Advertisement
Research Article

Thermal Tolerance of the Coffee Berry Borer Hypothenemus hampei: Predictions of Climate Change Impact on a Tropical Insect Pest

  • Juliana Jaramillo mail,

    jjaramillo@icipe.org

    Affiliations: Institute of Plant Diseases and Plant Protection, University of Hannover, Hannover, Germany, International Center of Insect Physiology and Ecology (icipe), Nairobi, Kenya

    X
  • Adenirin Chabi-Olaye,

    Affiliation: International Center of Insect Physiology and Ecology (icipe), Nairobi, Kenya

    X
  • Charles Kamonjo,

    Affiliation: International Center of Insect Physiology and Ecology (icipe), Nairobi, Kenya

    X
  • Alvaro Jaramillo,

    Affiliation: Centro Nacional de Investigaciones de Café, Manizales, Colombia

    X
  • Fernando E. Vega,

    Affiliation: Sustainable Perennial Crops Laboratory, United States Department of Agriculture, Agricultural Research Service, Beltsville, Maryland, United States of America

    X
  • Hans-Michael Poehling,

    Affiliation: Institute of Plant Diseases and Plant Protection, University of Hannover, Hannover, Germany

    X
  • Christian Borgemeister

    Affiliation: International Center of Insect Physiology and Ecology (icipe), Nairobi, Kenya

    X
  • Published: August 03, 2009
  • DOI: 10.1371/journal.pone.0006487

Abstract

Coffee is predicted to be severely affected by climate change. We determined the thermal tolerance of the coffee berry borer , Hypothenemus hampei, the most devastating pest of coffee worldwide, and make inferences on the possible effects of climate change using climatic data from Colombia, Kenya, Tanzania, and Ethiopia. For this, the effect of eight temperature regimes (15, 20, 23, 25, 27, 30, 33 and 35°C) on the bionomics of H. hampei was studied. Successful egg to adult development occurred between 20–30°C. Using linear regression and a modified Logan model, the lower and upper thresholds for development were estimated at 14.9 and 32°C, respectively. In Kenya and Colombia, the number of pest generations per year was considerably and positively correlated with the warming tolerance. Analysing 32 years of climatic data from Jimma (Ethiopia) revealed that before 1984 it was too cold for H. hampei to complete even one generation per year, but thereafter, because of rising temperatures in the area, 1–2 generations per year/coffee season could be completed. Calculated data on warming tolerance and thermal safety margins of H. hampei for the three East African locations showed considerably high variability compared to the Colombian site. The model indicates that for every 1°C rise in thermal optimum (Topt.), the maximum intrinsic rate of increase (rmax) will increase by an average of 8.5%. The effects of climate change on the further range of H. hampei distribution and possible adaption strategies are discussed. Abstracts in Spanish and French are provided as supplementary material Abstract S1 and Abstract S2.

Introduction

The impact of climate change on natural systems has emerged as one of the most critical issues faced by humankind. According to the Intergovernmental Panel on Climate Change (IPCC) [1], an increase in the mean global temperature of 1.4° to 5.8° C is expected by the end of the 21st century [2]. IPCC [1] provides an overview of our scientific understanding on climate change, and this assessment offers evidence of impact on, among others, natural biological systems [1]. Global climate change is likely to directly influence the dynamics of all trophic levels and further disrupt the multitrophic interactions among the different communities [3], [4].

In addition, climate change represents an immediate and unprecedented threat to agriculture. A 10–20% decline in overall global crop yields is predicted by 2050 [1]. This is of particular importance for crops such as coffee, which serves as the economic foundation for many countries in the tropics, and on which millions of people depend for their subsistence. Out of 103 species in the genus Coffea (Rubiaceae), only two are commercially traded: C. arabica L. and C. canephora Pierre ex A. Froehner [5], [6]. In terms of monetary value, coffee is the most heavily traded commodity in the world after oil [7]. Around 70% of the world's coffee is produced by small-scale farmers, with over 20 million coffee-farming families – equivalent to more than 100 million people - depending on its production for their subsistence [8]. Recent studies from Brazil, Mexico and Uganda show that even minimal increases in the mean temperature due to climate change will have disastrous consequences for coffee production, in some cases reducing the area presently suitable for coffee production by up to 95% [9][11]. Coffea canephora (widely known as robusta coffee) is native to humid forests or the lowland forests of the Congo River Basin, an area with elevations ranging from 0–1,200 meters above sea level (m.a.s.l.) [5], and an average temperature of 24–26°C [12]. Coffea arabica, regarded as the highest quality coffee, is native to the highlands of South Western Ethiopia where it grows naturally as an understory tree in forests at elevations ranging from 1,600–2,800 m.a.s.l. [5], and an average temperature of 18–21°C [13]. Above or below these temperatures the yield and quality of C. arabica is greatly reduced [14], [15]. Of the total world coffee production, 60% is arabica coffee [16]. Since the vast majority of C. arabica and C. canephora are grown in the tropics they are especially vulnerable to global climate change [17]. Climate-induced stress may render plants more vulnerable to opportunistic herbivores [18]. Furthermore, the direct effects of temperature on herbivores are likely to be larger and more important than any other factor associated with climate change like drought, CO2 levels, etc. [19][21]. Effects of climate change on insect herbivores can be direct, through impacts on their life history traits and number of generations per year [22], phenology [23], winter mortality [24] and distribution range [25], or indirect, e.g., when host-parasitoid interactions are affected [26] or when insects respond to climate-induced changes on the host plant [27].

Knowledge on thermal tolerance is essential to predict the effects of climate change in an organism [28]. Such information has never been used to predict the effects of climate change, i.e., global warming on the coffee berry borer Hypothenemus hampei Ferrari (Coleoptera: Curculionidae: Scolytinae) the most important pest of coffee throughout the world [29], [30] (Fig. 1). In this paper, we determine the thermal tolerance for H. hampei and make some inferences on the effects of climate change on the pest and on coffee production in the tropics, using climatic data from four coffee producing areas in Africa and South America. In addition, the original host plant and the possible area of origin of H. hampei as well as the reasons for the absence of the H. hampei in the presumed area of origin of C. arabica are discussed.

thumbnail

Figure 1. Female of the coffee berry borer (a), and female Hypothenemus hampei penetrating a coffee berry (b); (Photos: (a) Eric Erbe (USDA, ARS); (b) Gonzalo Hoyos CENICAFE).

doi:10.1371/journal.pone.0006487.g001

Results

Effect of temperature on the colonization of coffee berries and mortality of H. hampei colonizing females

The proportion of colonizing H. hampei females in the different positions inside the berries and their mortality/survival as a function of temperature are presented in Fig. 2. Across temperatures, <25% of colonizing females failed to penetrate the berries (position A) (Fig. 2a). In general the Logistic Model gave a good fit to the data for both the position of the colonizing females inside the berries (χ142 = 2662.19, P<0.0001) and mortality cases (χ72 = 211.87, P<0.0001).

thumbnail

Figure 2. Hypothenemus hampei colonizing females, % in positions inside the berry (a); % mortality or failure to penetrate the coffee berry (b).

doi:10.1371/journal.pone.0006487.g002

The position of the H. hampei females in the berries was significantly affected by temperature (χ72 = 953.92, (7, N = 5099) P<0.0001). The highest proportion of H. hampei females found in position D was recorded at 25°C (74.0%), followed by 23°C (54.2%), 20°C (52.3%) and 33°C (49.7%) (Fig. 2). At 15°C and 35°C there was no oviposition and the proportion of females found in the position B and C were 19.8%, 76.9% and 93.3%, 3.0%, respectively, indicating that at 15°C the colonizing females reached the endosperm but did not oviposit. On the other hand at 35°C the colonizing females did not reach the endosperm and remained in position B (Fig. 2a). Likewise, the temperature significantly affected the mortality/survival (χ72 = 546.15, (7, N = 5099) P<0.0001). The highest numbers of live females were found at 15°C (93.4%), followed by 25°C (83.8%) and 20°C (75.4%) (Fig. 2b). The highest numbers of dead H. hampei females were recorded at 35°C followed by 33°C (41.9% and 26.0%, respectively). In general, the proportion of surviving H. hampei females was high from 15 to 25°C; at higher temperatures (27 and 30°C), survival started to decrease considerably (Fig. 2b).

Effect of temperature on the developmental rate of H. hampei

None of the H. hampei life stages developed successfully at 15 and 35°C. The youngest life stages (egg and L1) developed between 20–33°C, whereas second instar larvae, prepupa, pupa and adult developed only between 20–30°C (Table 1). For all H. hampei life stages, the development time decreased significantly with temperature (between 20–30°C for egg, larvae 1 and 2; and 20–27°C for later life stages). At 33°C females oviposited but subsequent dissections revealed that 95% of the L1 died after eclosion. The developmental time of H. hampei immature stages was significantly influenced by temperature (Table 1). The duration of all immature stages except L1 (F4, 15 = 4.92, P = 0.0161), was significantly longer at 20°C than at 23, 25, 27 and 30°C, i.e. for egg (F4, 15 = 29.51, P<0.0001), L2 (F4, 15 = 39.0, P<0.0001), pre-pupa (F4, 15 = 8.65, P = 0.0021), pupa (F4, 15 = 22.40, P<0.0001). Egg to adult developmental time differed significantly at all temperatures tested (F4, 15 = 305.88, P<0.0001) (Table 1).

thumbnail

Table 1. Mean (±se) developmental time (in days) of Hypothenemus hampei life stages at different temperatures.

doi:10.1371/journal.pone.0006487.t001

For all beetle life stages, significant relationships between the developmental rate and temperatures were recorded (Table 2). In egg, pre-pupa, pupa and egg to adult time, the relationships were strongly linear, whereas a weaker relationship was recorded for L1, and linear regressions did not yield a good fit for development of the L2 (Table 2). For egg to adult, the lower developmental threshold was 14.9°C and the thermal requirement for completion of the pre-reproductive phase was calculated as 262.47 degree-days above the lower developmental threshold (Table 2).

thumbnail

Table 2. Estimates of the linear regression analyses (N = 15), lower thermal thresholds and the thermal constants for Hypothenemus hampei life stages.

doi:10.1371/journal.pone.0006487.t002

Developmental rates increased linearly between 15 and 27°C for prepupa, pupa and adult, and between 15 and 30°C for eggs and L1 (Fig. 3). In general, the non-linear model gave a good fit to the data sets within a range of 20–27°C for L1, pre-pupae, pupae and egg to adult development, and between 20–30°C for eggs and L1 (Fig. 3).

thumbnail

Figure 3. Effect of temperature on the developmental rates of Hypothenemus hampei.

doi:10.1371/journal.pone.0006487.g003

The fitted parameters of the model are presented in Table 3. Based on the non-linear models, the optimum temperature for the development of H. hampei egg and L1 was estimated as 30–32°C and for L2, pre-pupa, pupa and adult between 27–30°C. The lower and upper developmental threshold for all life stages was estimated as 14.9 and 32°C, respectively (Table 2 and Fig. 3).

thumbnail

Table 3. Fitted parameters of the non-linear modified Logan model [84], [85] for Hypothenemus hampei life stages.

doi:10.1371/journal.pone.0006487.t003

Life table parameters of H. hampei

Life table parameters are presented in Table 4. The intrinsic rate of increase (rm) was significantly higher at 25 and 27°C. Similarly, the reproductive rate (R0) significantly differed among temperatures tested, and was highest at 25°C followed by 27°C. The lowest reproductive rate was recorded at 30°C. With 68.0 and 11.8 days, the maximum generation time (G) and the doubling time (t), respectively, were obtained at 20°C. The finite rate of increase (λ) remained almost constant at all temperatures tested (Table 4).

thumbnail

Table 4. Average (+ se) population growth parameters of Hypothenemus hampei at five constant temperatures.

doi:10.1371/journal.pone.0006487.t004

Effect of temperature on fecundity of H. hampei

Pre-oviposition period and total H. hampei fecundity were significantly affected by temperature (F4, 14 = 8.08, P = 0.0035) and (F4, 14 = 40.97, P<0.0001), respectively. The longest pre-oviposition periods were recorded at 20°C and 23°C. Total fecundity measured as number of eggs laid by colonizing females before oviposition of the F1 females was significantly higher at 20°C (296.9 eggs) and lowest at 30°C (64.3 eggs) (Table 5). No differences were recorded in sex ratio, which ranged from 0.84 to 0.9 for all temperatures tested.

thumbnail

Table 5. Mean (±se) of pre-oviposition period, total fecundity, daily fecundity and sex ratio of Hypothenemus hampei at constant temperatures.

doi:10.1371/journal.pone.0006487.t005

Estimated number of H. hampei generations, warming tolerance and thermal safety margins in four coffee growing locations in East Africa and South America

Based on the number of degree-days in the four sites in Colombia, Kenya, Ethiopia and Tanzania (Table S1), the estimated number of generations of H. hampei per year ranged from 0.0 to 4.71. The calculated number of H. hampei generations for Kilimanjaro (Tanzania) and Chinchiná (Colombia) were very similar, ranging from 2.39 to 4.71, and 2.95 and 4.30, respectively. The number of H. hampei generation per year ranged between 2.03–3.13 for Kisii, Kenya (Fig. 4). With 0.0–2.02 the lowest number of beetle generations per year was estimated for Jimma, Ethiopia (Fig. 4).

thumbnail

Figure 4. Generations of Hypothenemus hampei in study sites in Colombia, Ethiopia, Tanzania, and Kenya.

doi:10.1371/journal.pone.0006487.g004

The Gaussian times a Gompertz model gave a good fit to the rm vs. temperature data sets within the range of 10–33°C (r2 = 0.98, F = 1592.28 P<0.0001; Fig. 5). The fitted parameters of the model were estimated as rmax = 0.14 (se = 0.003), ρ = 1.17 (se = 0.496), Topt = 26.7 (se = 0.421), Δ = 0.02 (se = 0.004). These analyses suggest that the rm is strongly influenced by temperature.

Calculated data on warming tolerance (WT) and thermal safety margins (TSM) of the insect for the three East African locations showed considerably high variability compared to the Colombian site (Table S2). The two variables were strongly and positively related (r = −0.936, P = 0.0001). However, the intrinsic rate of increase was significantly negatively correlated with the TSM for the period between 1975 and 2004. For every 1°C rise in Topt., rmax will increase by an average of 8.5%.

Discussion

Global warming is already affecting the bionomics of arthropods. Detailed studies from the temperate zones on several species report mainly positive effects on insect fitness and distribution range [31], [19], [22], [32]. On the other hand, surprisingly, the potential impact of global warming in tropical insects has only been studied in hematophagous insects such as malaria-transmitting Anopheles spp. (Diptera: Culicidae) [33] and the tsetse fly Glossina pallidipes Austen (Diptera: Glossinidae) [34]. In this paper, we report the first detailed analysis on the potential effects of climate change on the bionomics of H. hampei, the most devastating pest of coffee throughout the world. In a recent paper Deutsch et al. [28] predicted losses in tropical ectotherm biodiversity due to global warming.

In the absence of available long-term abundance data of H. hampei from the field in coffee producing areas of Africa and the Americas, we resorted to detailed laboratory studies on the effect of temperatures on the bionomics of the beetle. Previous studies on the basic biology of the insect were either carried out in the field [35][37], or in the laboratory with only one temperature regime [e.g.], [ 38][40]. Using a recently developed experimental protocol [41] we were able to study the effects of eight different temperature regimes on the bionomics of H. hampei. According to Stevens' [42] climatic variability hypothesis, the thermal tolerance of an insect is directly proportional to the climatic variability the organism is exposed to. Our study shows that the prediction of Steven's hypothesis was supported by the case of H. hampei. The estimated extremes for H. hampei survival are 15 and 32°C. Although our model predicts fastest development of the insect between 27–30°C, there is clear trade-off between development time and reproductive success as previously shown for other insects [43]. The highest rate of survival in colonizing females was recorded at the lowest temperature tested (Table 1).

H. hampei attacks and successfully develops on both C. arabica and C. canephora [44]. Moreover, the coffee berry borer is one of the few herbivores feeding on the endosperm of coffee due to its ability to detoxify caffeine [45], [46]. Coffea arabica is believed to have originated in the south-western highlands of Ethiopia, south-east Sudan (Boma Plateau) and the area around Mt. Marsabit in Kenya where it naturally grows as an understory tree in forests [5]. In this area, mean air temperatures range between 18–22°C [13]. In contrast, C. canephora is native to a vast area that covers West and West-Central Africa (Cameroon, Congo, Central African Republic, Democratic Republic of Congo, Gabon), North-East and East Africa (Sudan, Tanzania, Uganda), and Southern Africa (Angola). Yet, the exact limit of the centre of origin is difficult to determine because of introduction and naturalization [5]. In this area, mean annual air temperatures are higher than in the Ethiopian highlands [12]. For many years, there has been controversy in the literature about the geographic origin of the pest [47][49] and its original host plant(s) [50], [51]. Resolving this mystery might have important implications for future breeding for host plant resistance in coffee, as well as better targeted explorations for natural enemies of the coffee berry borer. Based on our estimates on the thermal tolerance of H. hampei it is unlikely that the beetle is endemic to the area around Jimma (Ethiopia) due to the low annual minimum temperatures prevalent there. During an extensive survey, Davidson [51] did not find H. hampei in Ethiopia and Damon [29] later speculated that the absence of the pest is due to either specialized natural enemies, resistant varieties of C. arabica or exceptionally clean harvest practices in Ethiopian plantations. However, our analysis of 32 years of climatic data from Jimma indicate that before 1984 it was too cold for the insect to complete even one generation per year, but thereafter, because of rising temperatures in the area, the pest is now able to complete 1–2 generations per year/coffee season. This may explain why in a more recent study Mendesil et al. [52] reported widespread occurrence of the coffee berry borer in southwestern Ethiopia.

Campbell et al. [53] emphasized the usefulness of the lower threshold of development and the thermal constant of an insect to elucidate its potential distribution. Similar to C. arabica, C. canephora is an understory tree of lowland forests. Climatological data from shaded coffee plantations in Central America [54], [55] and East Africa [56] indicate a reduction in temperature between 2–6°C depending on the region, when compared to coffee grown without shade. Considering these findings with our data on thermal tolerance of H. hampei and the annual mean temperatures in the area of origin of C. canephora [5], we hypothesize that the original host plant of H. hampei is likely to be robusta coffee. Since the two coffee species naturally occur in relative proximity in Central and Eastern Africa [5] it could have been possible that the coffee berry borer actively or passively dispersed from its original host plant C. canephora to the closely related C. arabica. Moreover, in life table studies in the laboratory H. hampei performed significantly better on robusta compared to arabica coffee (J. Jaramillo, unpublished data).

How do the findings reported in this paper relate to potential climate change effects on the coffee berry borer? Bale et al. [19] suggest that direct effects of temperature are likely to be stronger and more important than other factors related to climate change such as CO2 levels, rainfall pattern, etc. Moreover, data on thermal tolerance of an insect is crucial to predict the possible effects of climate change [e.g.], [57], [21], [22], [58]. In the case of H. hampei, average daily temperatures >26°C could lead to a reduction of the maximum intrinsic rate of increase, and, consequently, reduced pest activity in coffee plantations. Over the last three decades, the average daily temperature per year ranged between 17.3–22.3°C for Ethiopia, 18.7–24.5°C for Kenya, 22.3–29.8°C for Tanzania and for Colombia 15.5–29.3°C (data from 1989 to 2007, as H. hampei was introduced in 1988 into the country). The potential number of H. hampei generations per year was in average 3.4 for Colombia, 3.1 for Kenya, 3.1 for Tanzania and 1.3 for Ethiopia. According to our predictive model, in regions where the actual average daily temperature has not yet reached 26.7°C, every 1°C increase, would also increase the actual r towards the maximum value by an average of 8.5%. Since population growth of H. hampei is exponentially related to temperature (see Fig. 5), an increase of this magnitude may profoundly influence the pest population dynamics. Thus, areas with higher seasonal temperature close to the optimum development of H. hampei would experience high pest pressure as indicated by the higher number of generations of pests during its active period of the year. Because the coffee berry is a finite resource, this could lead to increased borer dispersal, as more females may be competing for oviposition sites. Such a scenario could have devastating effects in coffee growing areas like Colombia where, because of well distributed precipitation leading to multiple flowering of the plants, there is yearlong supply of coffee berries [59]. The problem would be less severe in East Africa where there is a marked and prolonged dry period, with consequent absence of berries in the field for extended periods.

thumbnail

Figure 5. Hypothenemus hampei intrinsic rate of increase (rm) estimated as function of temperature (°C) using a Gaussian times a Gompertz model.

doi:10.1371/journal.pone.0006487.g005

According to Hodkinson [60] the eco-physiology of both insect and plants will predict the future distribution of insect pests when both host plant and herbivore are in close synchrony. Thus, in the case of a highly specialized herbivore like H. hampei, the effects of climate change on the insect and the plant cannot be separated. Assad et al. [10] for Brazil, Gay et al. [11] for Mexico and Grid [9] for Uganda predict that even a small increase in temperature due to climate change, will have serious consequences for coffee production in these countries, in some cases rendering production very difficult. Under a climate change scenario, species like H. hampei, whose distribution are restricted by both temperature and the availability of their host plants, will follow plant distribution [61]. Coffee is mainly grown in the tropics from 20–25°N to 24°S [61]. Reactions of insects to climate change include, among others, acclimation [62], [63], and changes in latitudinal and altitudinal distribution [64], [65], [66], [26]. A latitudinal expansion in C. arabica and C. canephora is problematic because both species are highly susceptible to changes in photoperiod, with effects ranging from a marked reduction of their growth phase to an inhibition of flower development [67]. Yet to date coffee is grown in, among others, Nepal (http://www.plantecnepal.com/) and the Yunnan province of China (http://www.yunnancoffee.org/), both areas outside the before mentioned tropical distribution range of C. arabica and C. canephora. Our data on thermal tolerance of H. hampei on the other hand would predict that the pest is well capable to thrive also under sub-tropical conditions. An altitudinal expansion as a coping strategy in a climate change environment is potentially feasible, though there are few areas in the tropics where coffee production could expand in altitude (e.g. the Kilimanjaro area of Tanzania, Mt. Kenya, and the mountain ranges of Colombia), considering that other requirements for successful coffee production like soil type and appropriate rainfall patterns have to be met [6]. For instance, Jaramillo [68] estimated that for the coffee growing area in Colombia an increase in temperature of 1°C would require to move the plantations by 167m in altitude to maintain the same productivity and quality in arabica coffee. Recent data from Uganda and Indonesia show that the coffee berry borer has already expanded its altitudinal distribution range and is now attacking coffee plantations at sites as high as 1,864 m.a.s.l. (Dr. Africano Kangire, National Agricultural Research Organization, Uganda & Dr. Surip Mawardi, Indonesian Coffee and Cocoa Research Institute, pers. comm.)

A proven strategy to alleviate the potentially negative effects of climate, especially warmer temperatures, on coffee production is the introduction of shade trees in coffee plantations [67], [69]. Shade trees mitigate microclimatic extremes and can buffer coffee plants from microclimate variability [70], reduce high solar radiation and buffer detrimental diurnal changes in air temperature and humidity [55], [71], leading to a decrease in the temperature around the coffee berries by up to 4°C [71] under low altitude conditions (i.e., <700 m.a.s.l.), and by up to 2°C under mid to high altitude conditions (i.e., >1,100 m.a.s.l.) [56]. Moreover, Teodoro et al. [72] recently demonstrated that densities of the coffee berry borer were significantly lower in shaded versus un-shaded coffee plantations, possibly because shade coffee agro-ecosystems can serve as a refuge for beneficial arthropods (native and introduced), leading to higher levels of biological control of H. hampei [73], [74]. Finally despite lower yields of shaded compared to un-shaded coffee, the berry weight is higher and the quality of coffee produced under shade is better [75], with overall favorable economics for small-scale producers [76].

Our findings indicate that H. hampei can survive and reproduce within a broad temperature regime and that the potential number of generations as an indicator of the pest status varies profoundly with daily seasonal temperature. The lower number of generations found in Ethiopia is probably a result of the lower temperature prevalent in the sampling area. Thus we believe the most appropriate way for coffee production systems to cope with climate change is to come back to the origins of coffee as an understory tree in the forests of Africa.

Materials and Methods

Insects

Females of the coffee berry borer were obtained from an H. hampei stock culture established in July 2005 with beetle-infested coffee berries collected from an organic coffee plantation located in South Kisii (Gucha), Western Kenya (0° 45' 49.85'' S, 34° 43' 1.76'' E). The colony is maintained at the International Center of Insect Physiology and Ecology (icipe), Nairobi, Kenya, where the insects are reared on ca. 150 days old coffee berries (C. arabica var. Ruiru 11) kept at room temperature (25±1°C), 70% ±5% relative humidity [RH], and a 12:12 h (L:D) photoperiod. Infested berries were kept inside square plastic containers (40×40×20 cm) with perforated lids (55 mm diameter) covered with insect gauze. The bottom of each container was layered with a 1.5 cm mixture of plaster of Paris and activated charcoal to maintain humidity and prevent the desiccation of the berries and the insects [41].

Experimental setup and data assessment

The study was conducted at icipe laboratories. Organically produced coffee berries (C. arabica var. Ruiru 11) ca. 150 days old were collected from the Kiambu district of Central Kenya (1° 10' S, 36° 49' 60 E, 1,723 m.a.s.l.). Once in the laboratory, berries were surface sterilized using the following protocol [77]: the berries were washed with detergent for 15 min, rinsed with tap water, then dipped in a 2% sodium hypochlorite solution for 10 min, rinsed again with sterile distilled water, soaked in a 2% potassium sorbate solution and finally rinsed with sterile distilled water. Subsequently the excess of water was removed with a paper towel and the coffee berries were allowed to dry at room temperature. Afterwards, the berries were placed in round plastic containers (23 cm diameter×6.8 cm depth) (approx. 150 berries per container) and exposed to large numbers of H. hampei females from the stock culture. After 2 h of exposure, berries that had been bored by one female were selected and transferred individually into each well of a 12-well microtiter plate (Costar® 3526, Corning Inc., Corning, NY, USA). Each well (23 mm diameter; 20 mm deep) was filled with a 0.5 cm layer of a mixture of plaster of Paris and charcoal [41]. Twelve holes (15 mm dia), coinciding with the wells, were perforated in the lid of every multiwell plate and covered with mesh to allow aeration of the experimental units and to prevent escape of the beetles. The multiwell plates were then transferred to temperature controlled climate chambers (SANYO® MIR-553, Sanyo Electrical Ltd., Japan) set at eight different constant temperatures (15, 20, 23, 25, 27, 30, 33 and 35°C), 80 ±5% RH, and a 12:12 h (L: D) photoperiod. To keep up the humidity inside the experimental units, distilled sterile water was added to each well every two, three days or daily for multiwell plates kept at 20–30°C, 15°C and 33–35°C, respectively.

Numbers of live and dead H. hampei colonizing females as well as their position inside the berries (see below) and number of borer life stages (i.e., eggs, larvae, prepupae, pupae and adults) were assessed daily for periods between 30 and 60 days depending on the temperature.

Four different positions based on the insect location within the berry have been identified by Bustillo et al. [78] as follows: (A), when the female is starting to colonize a new berry but the penetration in the exocarp has not taken place; (B), when the female has penetrated the berry but has not yet reached the endosperm; (C), when the female has started to bore into the endosperm but not to oviposit; and (D), when the female has produced a gallery in the endosperm, and one or more of its immature stages are found inside the gallery.

The evaluations concluded when egg laying by the F2 generation was observed, i.e., between 30 and 60 days after the infestation of the berries depending on the temperature. The coffee berries were dissected under a 10X stereomicroscope and the position of the colonizing female inside the berry was recorded and the numbers of H. hampei immature stages were counted. On a daily basis, five berries per temperature and per replicate were destructively sampled and dissected under the stereomicroscope. The experiment was repeated four times over time for insects kept at 15 and 25°C, and three times for insects kept at 20, 23, 27, 30, 33 and 35°C. A total of 5,925 berries/H. hampei colonizing females were used, divides as follows: 1,200 for each 15 and 25°C (evaluations for a period of 60 days); 900 for 20°C (evaluations for a period of 60 days); 600 for each 23 and 27°C (evaluations for a period of 40 days); 525 for 30°C (evaluations for a period of 35 days); and 450 for each 33 and 35°C (evaluations for a period of 30 days).

Climatic data and estimated number of generations of H. hampei at four locations in Africa and South America

Daily climatic data were obtained for four coffee growing areas, with three locations in East Africa (Jimma, Ethiopia; Kisii, Kenya; and Kilimanjaro, Tanzania) and one location in South America (Chinchiná - Colombia) (Fig. SI 1). Precipitation data were used to estimate the yearly blossoming period of the main coffee harvest in the different locations. A single heavy rain (>10 mm rain), followed by a prolonged dry period usually triggers the blossoming of a coffee tree [79]. The physiological development of a coffee berry from flowering to the harvest of the ripe berry takes around 32 weeks or 240 days [80]. H. hampei females start to search for suitable coffee berries around 100 days after flowering and oviposit inside the berries usually 20 days later [37]. In the absence of H. hampei population dynamics data in all four locations, the findings of [79] and [37] were used to estimate the probable time of H. hampei oviposition in the different locations in East Africa and Colombia. Therefore, long-term daily data on temperature (Fig. SI 1) and precipitation (data not shown) in the different locations only for the period between 120–240 days of coffee berry development, together with our laboratory derived data on degree-days for H. hampei, were used to estimate the number of potential H. hampei generations per year and location.

Statistical analysis

The mortality/survival and the positions of the colonizing H. hampei female inside the coffee berry (A, B, C and D) at each temperature were analyzed using logistic regression (PROC LOGISTIC) [81].

Differences in developmental times, survivorship and life history parameters between temperatures were analyzed by analysis of variance (ANOVA), using the general linear model (PROC GLM) [81]. An F test was used to test the significance of mean differences and mean values were separated using the Student Newman Keuls (SNK) test. The significance level was set at P≤0.05. Percentages were transformed to arcsine values before analysis. The significance level was set at P = 0.05, but back-transformed data are presented in the tables. For estimation of the lower developmental threshold (T0) which is the intercept over the slope of the regression, i.e., the numbers of day-degrees to complete the pre-reproductive phase and the thermal constant (Kc) which is defined as one over the slope, a regression over the linear range of the relationship between temperature (T) and developmental rates [R (T)] of the insect was used [53]. (1)

A modified Logan model [82] by Lactin et al. [83] was used to describe the relationship between temperature and development rate,(2)

Where e is the exponential function, T is the temperature in degrees Celsius (°C), ρ, Tmax, Δ and λ are fitted coefficients.

All parameters in nonlinear models were estimated by minimization of the sum of squared residuals. Parameters were tested against 0, based on non-overlap of 95% confidence intervals.

Life table statistics were calculated according to Hulting et al. [84] using SAS, with calculation of confidence intervals for all estimated parameters [85]. The two-sided t-tests values, as well as their respective P values were computed, and mean values were separated using a pairwise comparison between populations.

For Kilimanjaro (Tanzania), Jimma (Ethiopia), Kisii (Kenya), and Chinchiná (Colombia), the historical degree-days were calculated between 120 and 240 days after flowering. The number of degrees above the threshold degree-days, for a single day are calculated as follows:

Where Max. and Min. are daily maximum and minimum temperature (°C). If Min. temperature was lower than the minimum threshold T0 (see Table 2), then Min. temperature was set to minimum threshold. If Max. temperature was higher than the maximum threshold 33°C (see Table 2), then Max. temperature was set to 33°C. The estimated number of H. hampei generations per year was calculated by dividing historical cumulative degree-days per year and location by the experimental estimation of Kc.

Warming Tolerance and Thermal Safety Margins of H. hampei

Warming tolerance (average amount of environmental warming an ectotherm can tolerate before performance drops to fatal levels) and thermal safety margins (temperature at which the performance of the organism will start to decrease) were calculated according to Deutsch et al. [28] as follows:

and

Where CTmax is the critical thermal maximum of H. hampei, Topt is H. hampei thermal optimum and Thab is the current climatological temperature of the organism's habitat.

The intrinsic rate of increase Rm(T) was fitted to the climatic data of the four locations using a Gaussian times a Gompertz function to predict the physiological consequences of climate change on the fitness of H. hampei:

Where rmax and Topt are the predictions where the insect has a maximal fitness, ρ represents the increasing part of the population growth rate curve and Δ the declining part of the curve.

Supporting Information

Table S1.

Number of degree days available to the coffee berry borer (above 14.9C) in three locations in East Africa and one in South America.

doi:10.1371/journal.pone.0006487.s001

(0.08 MB DOC)

Table S2.

Warming Tolerance (WT) and Thermal Safety Margin (TSM) (calculated after Deutsch et al., 2008) for Hypothenemus hampei in three locations in East Africa and one in South America.

doi:10.1371/journal.pone.0006487.s002

(0.12 MB DOC)

Abstract S1.

Spanish Abstract - Resumen

doi:10.1371/journal.pone.0006487.s003

(0.03 MB DOC)

Abstract S2.

French abstract - resume

doi:10.1371/journal.pone.0006487.s004

(0.03 MB DOC)

Acknowledgments

We thank Mr. Isaac Gichia and Mr. Moses Mbugua from Mbumi Coffee Estate and Millers, Kiambu, for kindly providing the coffee to conduct the experiments. Eric Muchugu and Bruce Anani for providing us with the climatic data for the three African locations and the Agroclimatology Department of Cenicafé, Chinchiná Colombia for the climatic data of the Colombian site. Special thanks to Prof. Matthew Thomas (Penn State University), Dr. Matthew Ayres (Dartmouth College), Dr. Sean Rands (University of Bristol), Dr. Juan Barrera (ECOSUR, Mexico), and Dr. Tim Batchelor (University of Nottingham) for reviewing an earlier version of the manuscript.

Author Contributions

Conceived and designed the experiments: JJ. Performed the experiments: JJ CK. Analyzed the data: JJ ACO AJ. Contributed reagents/materials/analysis tools: AJ. Wrote the paper: JJ ACO FEV HMP CB.

References

  1. 1. IPCC (2007) Summary for Policymakers. Climate Change 2007, published for the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press. pp. 2–18. In.
  2. 2. Houghton JT, Ding Y, Griggs DJ, Noguer M, van der Linden PJ, et al. (2001) Climate Change 2001: The Scientific Basis. Cambridge: Cambridge University Press. 769 p.
  3. 3. Parmesan C, Yohe G (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature 421: 37–42.
  4. 4. van der Putten WH, et al. (2004) Trophic interactions in a changing world. Basic Appl Ecol 5: 487–494.
  5. 5. Davis AP, Govaerts R, Bridson DM, Stoffelen P (2006) An annotated taxonomic conspectus of the genus Coffea (Rubiaceae). Bot J Linn Soc 152: 465–512.
  6. 6. Vega FE, Ebert AW, Ming R (2008) Coffee germplasm resources, genomics, and breeding. Plant Breeding Rev 30: 415–447.
  7. 7. Vega FE (2008) The rise of coffee. Am Sci 96: 138–145.
  8. 8. Vega FE, Rosenquist E, Collins W (2003) Global project needed to tackle coffee crisis. Nature 435: 343.
  9. 9. GRID (2002) Vital Climate Graphics Africa. Global Resource Information Database, Arendal. Available at: http://www.grida.no/climate/vitalafrica/​english/23.htm.
  10. 10. Assad ED, Pinto HS, Zullo J, Helminsk AM (2004) Climatic changes impact in agroclimatic zoning of coffee in Brazil. Pesqui Agropecu Bras 39: 1057–1064.
  11. 11. Gay C, Estrada F, Conde C, Eakin H, Villers L (2006) Potential impacts of climate change on agriculture: A case of study of coffee production in Veracruz, Mexico. Climatic Change 79: 259–288.
  12. 12. Coste R (1993) Coffee: the plant and the product. London: Macmillan. 328 p.
  13. 13. Alègre C (1959) Climate and arabica coffee. Agron Trop. 14. : 23–58. (in French).
  14. 14. Pinto HS, Zullo Junior J, Assad ED, Brunini O, Alfonsi RR, et al. (2001) Climate risk zoning for coffee production in the São Paulo state (Brazil). Rev Bras Agrometeo. 9. : 495–500. (in Portuguese).
  15. 15. Damatta F, Cochicho-Ramalho JD (2006) Impacts of drought and temperature stress on coffee physiology and production: a review. Brazilian J Plant Physiol 18: 55–81.
  16. 16. International Coffee Organization (2008) Letter from the Executive Director. Coffee Market Report, July 2008. Available at: http://dev.ico.org/documents/cmr0708e.pd​f.
  17. 17. Addo-Bediako A, Chown SL, Gaston KJ (2000) Thermal tolerance, climatic variability and latitude. Proc R Soc London Ser B Biol 267: 735–745.
  18. 18. Cannon RJC (1998) The implications of predicted climate change for insect pests in the UK, with emphasis on non-indigenous species. Glob Change Biol 4: 785–796.
  19. 19. Bale J, Masters GJ, Hodkinson ID, Awmack C, Bezemer M, et al. (2002) Herbivory in global climate change research: direct effects of rising temperature on insect herbivores. Glob Change Biol 8: 1–16.
  20. 20. Kiritani K (2006) Predicting impacts of global warming on population dynamics and distribution of arthropods in Japan. Popu Ecol 48: 5–12.
  21. 21. Kiritani K (2007) Pest status of rice and fruit bugs (Heteroptera) in Japan. Glob Change Biol 13: 1586–1595.
  22. 22. Gomi T, Nagasaka M, Fukuda T, Higahara H (2007) Shifting of the life cycle and life history traits of the fall webworm in relation to climate change. Entomol Exp Appl 125: 179–184.
  23. 23. Dingemanse NJ, Kalkman VJ (2008) Changing temperature regimes have advanced the phenology of Odonata in the Netherlands. Ecol Entomol 33: 394–402.
  24. 24. Ayres MP, Lombardero MJ (2000) Assessing the consequences of climate change for forest herbivores and pathogens. Sci Total Environ 262: 263–286.
  25. 25. Karban R, Strauss SY (2004) Physiological tolerance, climate change amd a northward rage shift in the spittlebug Philaenus spuramis. Ecol Entomol 29: 251–254.
  26. 26. Menendez R, Gonzalez-Megias A, Lewis OT, Shaw MR, Thomas CD (2008) Escape from natural enemies during climate-driven range expansion: a case study. Ecol Entomol 33: 413–421.
  27. 27. Forkner RE, Marquis RJ, Lill JT, Corff J (2008) Timing is everything? Phenological synchrony and population variability in leaf-chewing herbivores of Quercus. Ecol Entomol 33: 276–285.
  28. 28. Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor , et al. (2008) Impacts of climate warming on terrestrial ectotherms across latitude. Proc Natl Acad Sci USA 105: 6668–6672.
  29. 29. Damon A (2000) A review of the biology and control of the coffee berry borer Hypothenemus hampei (Coleoptera: Scolytidae). Bull Entomol Res 90: 453–465.
  30. 30. Jaramillo J, Borgemeister C, Baker PS (2006) Coffee berry borer Hypothenemus hampei (Coleoptera: Curculionidae): searching for sustainable control strategies. Bull Entomol Res 96: 223–233.
  31. 31. Parmesan C, Ryrholm N, Stefanescu C, Hill JK, Thomas CD, et al. (1999) Poleward shifts in geographic ranges of butterfly species associated with regional warming. Nature 399: 579–583.
  32. 32. Musolin DL (2007) Insects in a warmer world: ecological, physiological and life-history responses of true bugs (Heteroptera) to climate change. Glob Change Biol 13: 1565–1585.
  33. 33. Patz JA, Olson SH (2006) Malaria risk and temperature: Influences from global climate change and local land use practices. Proc Natl Acad Sci USA 103: 5635–5636.
  34. 34. Terblanche JS, Clusella-Trullas S, Deer JA, Chown SL (2008) Thermal tolerance in a south-east African population if the tse-tse fly Glossina pallidipes (Diptera, Glossinidae): implications for forecasting climate change impacts. J Insect Physiol 54: 114–127.
  35. 35. Borbón-Martinez MO (1989) Bioecologie d'un ravageur des baies de caféier. Hypothenemus hampei Ferr. (Coleoptera: Scolytidae) et de ses parasitoides au Togo. Toulouse (France). Université Paul Sabatier de Toulouse. 185 p. PhD thesis.
  36. 36. Baker PS, Barrera JF, Rivas A (1992) Life history studies on the coffee berry borer (Hypothenemus hampei, Scolytidae) on coffee trees in southern Mexico. J Appl Ecol 29: 656–662.
  37. 37. Ruiz CR (1996) Effect of coffee fruit phenology on the life table parameters of the coffee berry borer Hypothenemus hampei (Ferrari). 87 p. BSc thesis, University of Caldas, Colombia. (in Spanish).
  38. 38. Romero JV, Cortina H (2007) Life tables of Hypothenemus hampei (Coleoptera: Curculionidae: Scolytinae) on three coffee accessions. Rev Col Entomol 33: 10–16.
  39. 39. Romero JV, Cortina H (2004) Fecundity and life cycle of Hypothenemus hampei (Coleoptera: Curculionidae: Scolytinae) on coffee introductions. Rev Cenicaf. 55. : 221–231. (in Spanish).
  40. 40. Fernandez S, Cordero J (2007) Biology of the coffee berry borer Hypothenemus hampei (Ferrari) (Coleoptera: Curculionidae: Scolytinae) under laboratory conditions. Bioagro 19: 35–40. (in Spanish).
  41. 41. Jaramillo J, Chabi-Olaye A, Poehling HM, Kamonjo C, Borgemeister C (2008) Development of a new laboratory production technique for coffee berry borer Hypothenemus hampei (Ferrari) (Coleoptera: Curculionidae, Scolytinae), using fresh coffee berries. Entomol Exp Appl 130: 275–281.
  42. 42. Stevens GC (1989) The latitudinal gradient and geographical range: how so many species coexist in the tropics. Am Nat 133: 240–250.
  43. 43. Roff DA (2000) Trade-offs between growth and reproduction: an analysis of the quantitative genetic evidence. J Evol Biol 13: 434–445.
  44. 44. Le Pelley RH (1968) Pests of coffee. London: Longmans, Green and Co. Ltd., London. 590 p.
  45. 45. Vega FE, Blackburn MB, Kurtzman CP, Dowd PF (2003) Identification of a coffee berry borer-associated yeast: does it break down caffeine? Entomol Exp Appl 107: 19–24.
  46. 46. Vega FE (2004) Coffee Berry Borer. In: Capinera JL, editor. Encyclopedia of Entomology, Vol 1,. Kluwer Academic Publishers. pp. 575–576.
  47. 47. Bergamin J (1943) Contribution to an improved knowledge of biology of the coffee berry borer Hypothenemus hampei (Ferrari 1867) (Col.: Ipidae). Arquiv Inst Biol. 14. : 31–72 (in Portuguese).
  48. 48. Ticheler J (1961) Etude analytique de l'épidémiologie du scolyte des graines de café Stephanoderis hampei Ferr. en Cote d'Ivoire. Meded Landbouw Wageningen 61: 1–49.
  49. 49. Benavides P, Vega FE, Romero-Severson J, Bustillo AE, Stuart JJ (2005) Biodiversity and biogeography of an important inbred pest of coffee, the coffee berry borer, Hypothenemus hampei (Ferrari). Ann Entomol Soc Am 98: 359–366.
  50. 50. Baker PS (1984) Some aspects of the behaviour of the coffee berry borer in relation to its control in southern Mexico (Coleoptera: Scolytidae). Folia Entomol Mex 61: 9–24.
  51. 51. Davidson A (1967) The occurrence of coffee berry borer Hypothenemus (Stephanoderis) hampei in Ethiopia. Cafeì 8: 1–3.
  52. 52. Mendesil E, Jembere B, Seyoum E (2003) Ocurrence of coffee berry borer Hypothenemus hampei (Ferrari) (Coleoptera: Scolytidae) on Coffea arabica L. in Ethiopia. Ethiopian J Biol Sci 2: 61–72.
  53. 53. Campbell A, Frazer BD, Gilbert N, Gutierrez AP, Mackauer M (1974) Temperature requirement of some aphids and their parasites. J Appl Ecol 11: 431–438.
  54. 54. Barradas VL, Fanjul L (1986) Microclimatic characterization of shaded and open-grown coffee (Coffea arabica L.) plantations in Mexico. Agr Forest Meteorol 38: 101–112.
  55. 55. Vaast P, Bertrand B, Perriot JJ, Guyot B, Genard M (2006) Fruit thinning and shade improve bean characteristics and beverage quality of coffee (Coffea arabica L.) under optimal conditions. J Sci Food Agr 86: 197–204.
  56. 56. Kirkpatrick TW (1935) The Climate and Eco-Climates of Coffee Plantations. East African Agricultural Research Station, Amani 66 p.
  57. 57. Hodkinson ID, Bird J, Miles JE, Bale JS, Lennon JJ (1999) Climatic signals in the life histories of insects: the distribution and abundance of heather psyllids (Strophingia spp.) in the UK. Funct Ecol 13: 83–95.
  58. 58. Calosi P, Bilton DT, Spicer JI, Atfield A (2008) Thermal tolerance and geographical range size in the Agabus brunneus group of European diving beetles (Coleoptera: Dytiscidae). J Biogeogr 35: 295–305.
  59. 59. Arcila J, Jaramillo A, Baldion V, Bustillo AE (1993) Coffee flowering and its relationship to coffee berry borer control. Avance Técnico No 193. Colombia: CENICAFE, Chinchiná. 5 p. (in Spanish).
  60. 60. Hodkinson ID (1999) Species response to global environmental change or why ecophysiological models are important: a reply to Davis et al. J Anim Ecol 68: 1259–1262.
  61. 61. Ward NL, Masters GJ (2007) Linking climate change and species invasion: an illustration using insect herbivores. Glob Change Biol 13: 1605–1615.
  62. 62. Wellman FL (1961) Coffee: botany, cultivation, and utilization. London: L. Hill. 488 p.
  63. 63. Wilson RS, Franklin CE (2002) Testing the beneficial acclimation hypothesis. Trends Ecol Evol 17: 66–70.
  64. 64. Terblanche JS, Sinclair BJ, Klok CJ, McFarlane ML, Chown SL (2005) The effects of acclimation on thermal tolerance, desiccation resistance and metabolic rate in Chirodica chalcoptera (Coleoptera: Chrysomelidae). J Insect Physiol 51: 1013–1023.
  65. 65. Neuvonen S, Niemela P, Virtanen T (1999) Climatic change and insect outbreaks in boreal forests: the role of winter temperatures. Ecol Bull 47: 63–67.
  66. 66. Gaston KJ, Chown SL (1999) Elevation and climatic tolerance: a test using dung beetles. Oikos 86: 584–590.
  67. 67. Amaral do JAT, Rena AB, Amaral do JFT (2006) Seasonal vegetative growth of the coffee plant and its relationship with the photoperiod, fructification, stomatic resistance and photosynthesis. Pesqui Agropecu Bras 41: 377–384.
  68. 68. Jaramillo A (2005) Andean climate and coffee in Colombia. Colombian National Federation of Coffee Growers. Centro Nacional de Investigaciones de Café – Cenicafé. Colombia: Chinchiná. 192 p. (in Spanish).
  69. 69. Lin BB (2007) Agroforestry management as an adaptive strategy against potential microlimate extremes in coffee agriculture. Agr Forest Meteorol 144: 85–94.
  70. 70. Beer J, Muschler R, Kass D, Somarriba E (1998) Shade management in coffee and cacao plantations. Agroforest Syst 38: 139–164.
  71. 71. Jaramillo A (2005) Solar radiation and rainfall distribution within coffee plantations (Coffea arabica L.) Rev Acad Col Ci Ex Fís Ntles (Colombia) 29: 371–382. (in Spanish).
  72. 72. Teodoro A, Klein AM, Tscharntke T (2008) Environmentally mediated coffee pest densities in relation to agroforestry management, using hierarchical partitioning analyses. Agr Ecosyst Environ 125: 120–126.
  73. 73. Perfecto I, Rice RA, Greenberg R, van der Voort ME (1996) Shade coffee: a disappearing refuge for biodiversity. BioScience 46: 598–608.
  74. 74. Tylianakis JM, Klein AM, Tscharntke T (2005) Spatio-temporal variation in the effects of a tropical habitat gradient on Hymenoptera diversity. Ecology 86: 3296–3302.
  75. 75. Muschler RG (2001) Shade improves coffee quality in a suboptimal coffee zone of Costa Rica. Agroforest Syst 85: 131–139.
  76. 76. Gordon C, Manson R, Sundberg J, Cruz-Angon A (2007) Biodiversity, profitability and vegetation structure in a Mexican coffee agroecosystem. Agr Ecosyst Environ 118: 256–266.
  77. 77. Pérez J, Infante F, Vega FE (2005) Does the coffee berry borer (Coleoptera: Scolytidae) have mutualistic fungi? Ann Entomol Soc Am 98: 483–490.
  78. 78. Bustillo AE, Cardenas R, Villalba D, Benavides P, Orozco J, et al. (1998) Integrated pest management of the coffee berry borer Hypothenemus hampei (Ferrari) in Colombia.Chinchiná, Cenicafé. 134 p. (in Spanish).
  79. 79. De Alvim PT (1960) Moisture stress as a requirement for flowering of coffee. Science 132: 3423.
  80. 80. Salazar MR, Chaves B, Riaño N, Arcila J, Jaramillo RA (1993) Growth and development of the coffee berry Coffea arabica L. var. Colombia Rev Cenicafé 45: 41–50. (in Spanish).
  81. 81. SAS Institute 1999SAS/Stat user's guide. Cary, North Carolina: SAS Institute..
  82. 82. Logan JA, Wollkind DJ, Hoyt SC, Tanigoshi LK (1976) An analytical model for description of temperature dependent rate phenomenon in arthropods. Environ Entomol 5: 1133–1140.
  83. 83. Lactin DJ, Holliday NJ, Johnson DL, Craigen R (1995) Improved rate model of temperature-dependent development by arthropods. Environ Entomol 24: 68–75.
  84. 84. Hulting FI, Orr DB, Obrycki JJ (1990) A computer program for calculation and statistical comparison of intrinsic rates of increase and associated life table parameters. Fla Entomol 73: 601–612.
  85. 85. Maia A, de H, Luiz AJ, Campanhola C (2000) Statistical inference on associated fertility life table parameters using jackknife technique: computational aspects. J Econ Entomol 93: 511–51.