Author's personal copy Neotrop Entomol DOI 10.1007/s13744-014-0226-9
PEST MANAGEMENT
Population Dynamics and Temperature-Dependent Development of Chrysomphalus aonidum (L.) to Aid Sustainable Pest Management Decisions O C AMPOLO 1 , A M ALACRINÒ 1 , F L AUDANI 1 , V MAIONE2, L ZAPPALÀ3, V PALMERI1 Depto di Agraria, Univ “Mediterranea” of Reggio Calabria, Reggio Calabria, Italy Agenzia Regionale Sviluppo e Servizi in Agricoltura, Regione Calabria, Italy 3 Depto di Gestione dei Sistemi Agroalimentari e Ambientali, Univ of Catania, Catania, Italy 1
2
Keywords Briere, Florida red scale, Lactin, Logan, non-linear model, Taylor Correspondence V Palmeri, Depto di Agraria , Univ “Mediterranea” of Reggio Calabria, Reggio Calabria, Italy;
[email protected] Edited by Wesley AC Godoy – ESALQ/USP Received 29 January 2014 and accepted 28 May 2014 * Sociedade Entomológica do Brasil 2014
Abstract The increasing worldwide trades progressively led to decreased impact of natural barriers on wild species movement. The exotic scale Chrysomphalus aonidum (L.) (Hemiptera: Diaspididae), recently reported on citrus in southern Italy, may represent a new threat to Mediterranean citriculture. We studied C. aonidum population dynamics under field conditions and documented its development under various temperatures. To enable describing temperature-dependent development through the use of linear and non-linear models, low temperature thresholds and thermal constants for each developmental stage were estimated. Chrysomphalus aonidum was able to perform four generations on green parts (leaves, sprouts) of citrus trees and three on fruits. In addition, an overall higher population density was observed on samples collected in the southern part of the tree canopy. Temperature had a significant effect on the developmental rate; female needed 625 degree days (DD) to complete its development, while male needed 833 DD. The low threshold temperatures, together with data from population dynamics, demonstrated that C. aonidum is able to overwinter as second instar and as an adult. The results obtained, validated by those collected in the field, revealed few differences between predicted and observed dates of first occurrence of each C. aonidum instar in citrus orchards. Data on C. aonidum phenology and the definition of the thermal parameters (lower and upper threshold temperatures, optimum temperature, and the thermal constant) by nonlinear models could allow the estimation of the occurrence in the field of each life stage and would be helpful in developing effective integrated control strategies.
Introduction The globalization of human societies in recent years has caused a new design of worldwide flow of people and goods and led to the collapse of natural barriers to wild species movements (Liebhold & Tobin 2008). The movement of species, either intentional or accidental, increased with the intensity of human exchanges among different countries.
The accidental introduction of new species, in particular insects, does not often lead to important consequences for human activity, but sometimes non-indigenous species grow so fast in number that they could bring to important consequences to human health or to economic activity (Liebhold & Tobin 2008). In addition, exotic pests could cause environmental alterations and the displacement of endemic species (Lee 2002).
Author's personal copy Campolo et al
Many scale insects widely distributed in Italy are not indigenous but were introduced and then acclimatized, probably because of the particular position of the Italian peninsula in the Mediterranean basin for commercial exchanges and its climate, that allow the establishment of tropical and subtropical species (Longo et al 1995). One of the latest exotic species introduced in southern Italy was Unaspis yanonensis (Kuwana) (Hemiptera: Diaspididae), a diaspid scale that causes several damages to citrus groves, such as premature leaf and fruit drop (Campolo et al 2013). In the same way, the accidentally introduced Florida red scale, Chrysomphalus aonidum (L.), could represent a new threat to citrus crops (Palmeri et al 2011). Chrysomphalus aonidum is native to Asia and is currently widespread in tropical and subtropical regions of the world including North and South America, Africa, Australia, Pacific Islands, the Mediterranean, and the Far East. In Europe, C. aonidum has been found in Croatia, Cyprus, Greece, Malta, Romania, Serbia, Spain, Belgium, Bulgaria, Denmark, France, Germany, Great Britain, Hungary, Italy, Madeira, Netherlands, Poland, Portugal (where the species was eradicated), and Slovenia. Although in northern and central Europe the species has been recovered only in greenhouses, in southern Europe it is well established outdoors (Hlavjenková & Šefrová 2012). The Florida red scale is considered a serious pest of global economic importance (Beardsley & Gonzalez 1975). In particular, it has been recorded as a serious pest of citrus in Florida, Texas, Brazil, Mexico, Lebanon, Egypt, and Israel (Ben-Dov et al 2001). In some of these countries, the scale is effectively controlled by the parasitoids Aphytis holoxanthus DeBach and Pteroptrix smithi (Compere) (Hymenoptera: Aphelinidae) (Pellizzari & Vacante 2007), although their activity is often strongly disturbed by pesticide applications (Cohen et al 1988, Rehman et al 2000). The Florida red scale was ranked as one of the ten widely recovered Diaspididae on citrus in Europe (Pellizzari & Germain 2010) and is reported as an important pest of citrus in the Mediterranean basin (Gerson 2012). In Italy, this scale insect has been accidentally introduced several times, mainly through the international trade of ornamental plants, but it did not survive outside greenhouses or protective structures because of unsuitable climatic conditions (Longo et al 1994). Recent climate changes led to new opportunities for the establishment and spread of new pests (Porter et al 1991, Kocmánková et al 2010, Thomson et al 2010). Besides, the spread of pests may also be due to modifications in the control strategies and in the agronomic practices in general as well as to alterations in the natural enemies complex (Rehman et al 2000). Indeed, C. aonidium presence was recorded in Calabria (southern Italy) on citrus trees during 2007 (Pellizzari & Vacante 2007). The same phenomenon happened in Spain, where the species was first recovered on open field citrus in 2000
(Garcia Mari et al 2000). Chrysomphalus aonidum is a highly polyphagous species, including in its host range 192 plant genera belonging to 77 unrelated families (Hlavjenková & Šefrová 2012), with a clear preference for the genera Citrus, Musa, and Eucalyptus (Schweig & Grunberg 1936, Quayle 1941). Florida red scale grows on leaves, green twigs, and fruits, causing leaves yellowing, premature leaf and fruit drop, and stem dieback (Watson 2005). The control of C. aonidum, such as that of armored scales in general, is made harder by the presence of the scale cover almost all along the insect life cycle. The control of key armored scale pests is mainly based on chemical treatments both on cultivated and ornamental plants (Walker et al 1991, Frank 2012). Therefore, in order to maximize the treatment efficacy and keep pest populations below economically damaging levels, pesticide applications should be performed at the pest’s point of highest susceptibility. This is represented by the crawler stage which may be killed by contact with or by traversing the toxic residue. In order to identify this developing phase, the pest should be constantly monitored. To this end, acquiring knowledge on the pest biology and particularly on the relation between temperature and the pest life cycle is a very useful tool to establish an appropriate spraying time. This is even more important in the framework of a reduction of pesticide applications as enforced by the EU Directive on sustainable use of pesticides (Directive 2009/128/EC). Several studies on life history traits and their relevance for pest control have been conducted for key armored pests, such as the California red scale, Aonidiella aurantii (Maskell) (Yu & Luck 1988, Ewing et al 2002), or the Arrowhead scale, U. yanonensis (Kim & Kim 2013), but very few data are available for C. aonidum. Besides, although data on the temperature-dependent development of C. aonidum were already presented by Andrade et al (2008), the approach used in their paper did not allow a thorough understanding of the relationship with thermal values and data obtained could not be applied to field conditions. Furthermore, in this previous work, values of thermal parameters (thermal constant, lower and upper temperature thresholds, optimum temperature) were not provided. The aim of this work was to fill the gaps of knowledge about the role of temperature on the development of C. aonidum. Studies of C. aonidum field population dynamics and its development under controlled temperatures were carried out. A non-linear relationship was provided, and the thermal parameters were validated with field data, providing new information on the biology and physiology of this pest that could contribute to the development of IPM programs in the Mediterranean basin.
Author's personal copy C. aonidum Population Dynamics and Developmental Models
Material and Methods Field population dynamics Studies on the population dynamics of C. aonidum were carried out between the beginning of May 2008 and the end of April 2009. Citrus leaves and fruits were collected from an unsprayed sweet orange, Citrus sinensis, var Navelina organic orchard located in Locri, Reggio Calabria (38° 12′ 54.7″, 16° 14′ 40.6″), southern Italy. The trees in the experimental field were 15 years old, regularly pruned and irrigated. Samplings were carried out on a group of 16 infested trees located in the orchard. Every 7 days, four 1 to 2-year-old twigs of 40 cm and four fruits, one from each cardinal direction, were picked from each of four randomly selected trees. The collected samples were individually placed in a paper bag, sealed in a vinyl bag (40×50 cm), and placed in a refrigerator at 4°C until examined. Observations were carried out within 24 h from sampling. Ten leaves from each twig were randomly selected, and a circular area of 5.72 cm2 was randomly delimited on the leaf underside, which is generally more infested than the upper surface due to the preference of the species for spots not subjected to direct sunlight (Schweig & Grunberg 1936). A similar circular area was also observed on the surface of the sampled fruits. All the specimens of C. aonidum were examined and counted using a stereo-microscope (×7–90). They were classified as first instars, second instar males and females, and male pre-pupa and pupa; in addition, adult females were distinguished in pre-ovipositing and ovipositing, according to Hlavjenková & Šefrová (2012), observing the pygidial zone that becomes more round and closer to the rest of the body during oviposition. Air temperatures and relative humidity were recorded throughout the period of the study using a CR10 Measurement and Control Module, equipped with a CS215 temperature and relative humidity sensor (Campbell Scientific, Inc. Logan, UT USA) located in the experimental field.
Development under controlled temperatures In order to obtain C. aonidum specimens at mobile stage to evaluate the temperature-dependent development, scales collected in the field were reared at 25°C and 75% RH with 12:12 L/D photoperiod, using Cucurbita moschata as host. The experiment was conducted using green lemon fruits, previously cleaned thoroughly with distilled water. A 5-cm2 paper disk was applied on the lemon surface, and then the fruits were coated by immersion in paraffin previously melted at 50°C. After paraffin solidification, the paper disks were removed in order to ensure the establishment of the
scale in a circumscribed area, safeguarding the integrity of the fruits over time and reducing fruit transpiration. The developmental time was evaluated at eight temperatures (5, 10, 15, 20, 25, 30, 35, and 40°C), the relative humidity was kept at 75±5%, and the photoperiod was set at 12:12 L/D. A total of 30 newly hatched specimens of C. aonidum at the mobile stage (crawlers) were placed in the delimited space on each fruit surface prepared as earlier described and then placed in the climatic chamber for a total of ten fruits per treatment. Newly hatched crawlers were collected from the rearing from squashes bearing ovipositing adult females checked every 12 h. In order to obtain ten scales per fruit, excess specimens were randomly discarded after 36 h. Scales were unequivocally identified by taking a photograph and assigning a progressive number to each specimen. Every 24 h, the surface of each fruit was photographed by means of a camera (Olympus® XC50) coupled with a stereo-microscope (Olympus® SZX12); the developmental stage of each scale was recorded. This was done also in order to follow the development of first instar nymphs and to later determine their sex.
Data analysis Differences in the abundance of specimens from different directional exposures and the influence of temperatures on the developmental time were assessed by applying the Kruskal-Wallis non-parametric test because the data did not fulfill the assumption of normality (Kolmogorov-Smirnov test), followed by the Steel’s multiple comparison procedure. In order to analyze and to describe the relationship between the development of C. aonidum and the temperature, we adopted a linear and four non-linear models. Furthermore, the sine model proposed by Allen (1976) was used first to calculate the thermal accumulation from the 1st of January until the first appearance of crawlers of C. aonidum in the field and then to predict the first occurrence of each instar in the field. The linear model is often used to determine insect thermal constant (k) and low temperature threshold (L t ) (Kontodimas et al 2004, Prasad et al 2012). This threshold sets the temperature limit below which there is no measurable development, while k represents the amount of heat units (degree days—DD) above Lt needed to complete a developmental stage (Jalali et al 2010, Prasad et al 2012). The linear model is defined as follows: Rt ¼ a þ bT where developmental rate (Rt) is expressed as the reciprocal of the developmental period (days), T is the temperature, a is the intercept, and b is the slope. Fit lines were derived for each stage and sex; Lt and k values were computed by solving
Author's personal copy Campolo et al
the equations “−intercept/slope” and “1/slope” of the linear portion of Rt curves (Campbell et al 1974, Kim & Kim 2013). Thermal constant standard error (SE) was obtained following the procedure indicated by Campbell et al (1974): SEk ¼
SEb b2
Moreover, Briere-1, Taylor, Logan, and Lactin non-linear models were applied (Table 1). Indeed, the relationship between developmental rate and constant temperatures tends to be non-linear for arthropods, so the use of the degree days model (Allen 1976) that does not include the non-linearity tending to developmental thresholds may produce biased results. The non-linear models adopted included also parameters (temperature and upper, optimal, and lower thermal thresholds) that have an ecological meaning and that allow to calculate values, such as the upper thermal threshold (Tmax) and the optimal developmental temperature (Topt), that cannot be calculated through a linear regression. For each model, initial calculation parameters were based on values obtained from the linear model (Lt) or results from the other models (Tmax and Topt) and then iterated in order to perform a better fit of the data. In Logan and Lactin models, Topt was estimated by iteration of temperature values until reaching the maximum developmental rate. While in Logan model, it is not possible to estimate Lt parameter, as this equation approaches to zero asymptotically; in Lactin model, it was estimated by iterating the function in order to determine the lower point where it crosses the x-axis. The Briere model was calculated through the parameters Lt and Tmax, allowing the estimation of Topt using the criteria proposed by Briere et al (1999): pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 4T max þ 3Lt þ 16T 2max þ 9L2t −16Lt T max T opt ¼ 10 The estimated performance of each mathematical model was assessed by using the R2 and residual sum of squares (RSS) coefficients by Akaike information criterion (AIC) and
Bayesian information criterion (BIC) that integrate the information about the goodness of fit with the settings of the proposed models. Data obtained from these models, together with field thermal data, were entered into the Allen sine model (Allen 1976). The predicted dates were compared with dates of first appearance recorded in the field where these were calculated as midpoint between the date of first occurrence of a given biological instar and the previous sample date. Data analyses were carried out using R statistical software version 2.3.0 (R development Core Team 2008) and Microsoft ® Excel ® 2013.
Results Field population dynamics The data collected showed that the scale was able to perform four generations on leaves. The first oviposition was observed between May and June, followed by peaks at the end of July, at the end of September, and during the first days of November (Fig 1a). The third oviposition reached a higher intensity compared to the others, with 11.87±0.14 specimens×(cm2)−1 recorded. The number of first instars was null from the end of February until the first days of May. Females at second instar also showed four peaks, with a maximum during the first days of October (Fig 1b), reaching 3.31±0.09 specimens×(cm2)−1. In contrast, second instars male (Fig 1c) were not abundant during summer months, but their presence increased in September with a maximum of 1.64±0.34 scales×(cm2)−1. Specimens belonging to both sexes at second instar were present at low numbers during all winter. Adult females at the pre-ovipositing phase presented also four peaks of density during the study: the first was observed in May; afterward, their presence increased at the end of June, at mid-August, and between October and November, when a peak of 1.78±0.12 scales×(cm2)−1 was observed (Fig 1d). The population dynamics of ovipositing females
Table 1 Models applied to describe the temperature-dependent development of Chrysomphalus aonidum. Model
Equation
References
Briere-1 Taylor Logan Lactin
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Rt ¼ a temp ðtemp−Lt Þ T max −temp Rt =ρ×exp{−0.5×[(temp−Topt)/Lt]2} Rt =α×{exp(ρ×temp)−exp[ρTmax −(Tmax −temp)/Δ]} Rt =λ+{exp(ρ×temp)−exp[ρTmax −(Tmax −temp)/Δ]}
Briere et al (1999) Taylor (1981) Logan et al (1976) Lactin et al (1995)
Rt is the rate of development function of temperature (temp), Tmax is the upper thermal threshold, Topt is the optimal developmental temperature, Lt is the lower developmental threshold, Δ=Tmax −Topt, ρ is the developmental rate at Topt and a, and α and λ are fitting parameters.
Author's personal copy C. aonidum Population Dynamics and Developmental Models Fig 1 Field population dynamics of Chrysomphalus aonidum [mean live specimens×(cm2)−1] on leaves: a first instar; b second instar females; c second instar males; d pre-ovipositing females; e ovipositing females; f male prepupae; and g male pupae.
reflected that of pre-ovipositing females, delayed of approximately 2 weeks (Fig 1e). The presence of adult females was constant during all winter, with a mean number of 0.21± 0.009 specimens×(cm2)−1. The presence of Florida red scale males at pre-pupal and pupal stages was higher, respectively, during the first days of August [0.5±0.11 scales×(cm2)−1] and between September and October [0.4±0.08 scales×(cm2)−1]. Scales at these stages also overwintered with a low number of specimens. The directional exposure significantly influenced the number of C. aonidum specimens on leaves (χ2 =73.424; df=3; p≤ 0.001), with a significantly higher number at southern
exposure, where 0.64±0.01 scales×(cm2)−1 on average were found, while the other exposures did not show any significant difference, with 0.48± 0.01, 0.52±0.01, and 0.51 ± 0.01 at north, east, and west, respectively (Fig S1 in Supplementary Material). The population dynamics of C. aonidum was evaluated also on fruits. These observations showed that Florida red scale is able to perform three generations on fruits, the first in July, the second one in mid-August, and the third one between October and November (Fig 2). These data overlap with those recorded on leaves, although the number of scales was always higher on fruits than on leaves.
Author's personal copy Campolo et al
Fig 2 Field population dynamics of Chrysomphalus aonidum [mean live specimens×(cm2)−1] on fruits: a first instar; b second instar females; c second instar males; d pre-ovipositing females; e ovipositing females; f male prepupae; and g male pupae.
During the year of study, the minimum temperature was −1.3°C, recorded in mid-February, while the highest temperature (39.6°C) was recorded during the first days of August. The winter mean temperature was 8.9±0.24°C, the mean minimum was 4.09±0.29°C and the mean maximum was 14.8±0.4°C (Fig S2 in Supplementary Material). Development under controlled temperatures Temperature had a significant effect on the development time and, subsequently, on the developmental rate (χ2 =
179.53; df=4; P
