The reproductive performance of the Mupli beetle, Luprops tristis, in relation to leaf age of the para rubber tree, Hevea brasiliensis.

An analysis of host plant leaf age preferences and phenology studies led to the predictions that tender rubber plant leaves are essential for the completion of the life cycle of the Mupli beetle, Luprops tristis Fabricius (Coleoptera: Tenebrionidae) and that low tender leaf availability during the post-dormancy stage will limit the beetle population. Analyses of the effects of feeding the beetles leaves of various ages, nitrogen (N) content, and moisture content on fecundity and the duration of post-dormancy survival were carried out. The results showed that tender leaf availability during the post-dormancy phase of L. tristis is a critical factor that determines the survival of L. tristis adults and the subsequent generation. The control of powdery mildew ( Odium hevea) disease-mediated premature leaf fall in rubber plantations may regulate the beetle population. A peak in fecundity during the early phase of post-dormancy is proposed as an adaptive mechanism of L. tristis to synchronize egg production and feeding with tender leaf availability in rubber plantations. Variations in nutrient levels and moisture content between deciduous rubber tree leaves of different ages are attributed to the leaf nutrient resorption mechanism of senescing leaves. These results established that tender leaves with high N and moisture levels are essential for post-dormancy survival and that N influences fecundity. The results of the experiments could aid decision making regarding the population management and control of L. tristis in rubber plantations. This is an open access paper. We use the Creative Commons Attribution 3.0 license that permits unrestricted use, provided that the paper is properly attributed.

Introduction

Massive seasonal invasions of the Mupli beetle, Luprops n class="Species">tristis (Fabricius) (Coleoptera: Tenpan>ebrionpan>idae), cause various problems. These beetles enpan>ter residenpan>tial buildinpan>gs followinpan>g summer showers, are nocturnpan>al, are attracted to light, produce pan> class="Disease">allergenic defensive secretions, and go dormant for 8–9 months, making them an extreme nuisance in the rubber plantation belts of southern India ( Sabu et al. 2008 ; Sabu and Vinod 2009a , b). Their very high abundance, concealment in rubber plantation litter layers, aggregation in residential buildings, and lack of natural enemies ( Aswathi and Sabu 2011 ) make controlling them with conventional methods unfeasible. An analysis of their host plant preferences and habits revealed that plantation litter stands of rubber trees, Hevea brasiliensis (Willdenow ex Adrien De Jussieu) Müller Argoviensis 1865 (Malpighiales: Euphorbiaceae), are the breeding and feeding habitats, and prematurely fallen tender rubber leaves are the preferred food source ( Sabu and Vinod 2009a , b; Sabu et al. 2012 ). The breeding phase of the post-dormancy beetles is perfectly synchronized with the annual leaf shedding and sprouting of new rubber plant leaves during the pre-summer period ( Sabu and Vinod 2009b ), and beetles of certain developmental stages (eggs, larval instars, pupae, and teneral adults) peak at the premature fall of the tender leaves (Vinod and Sabu 2009). These findings indicate that tender leaves are important for the completion of the life cycle of L. tristis and that the control of tender leaf availability will limit the beetle population ( Sabu and Vinod 2009b ). An empirical analysis of how the lack of tender leaves will affect the reproductive performance and survival of post-dormancy beetles was undertaken as the primary objective of the present study.

The observation that n class="Species">L. tristis is attracted to anpan>d prefers to feed onpan> tenpan>der leaves necessitates anpan> anpan>alysis of the nutritionpan>al quality of leaves of various ages. There is widespread evidenpan>ce from herbivorous inpan>sects that age-related variationpan> inpan> leaf nutrienpan>t quality, especially inpan> pan> class="Chemical">nitrogen (N) and moisture levels, affects insect performance and that tender leaf availability is a major factor in determining the most suitable periods for larval development and for the optimal reproductive capabilities of adults ( Schroeder 1986 ; Awmack and Leather 2002; Haukioja 2003 ; Riipi et al. 2004 ; Ruusila et al. 2005 ; Van Asch and Visser 2007 ). However, no data exist on the age-related variations of rubber leaf quality. Hence, age-related variations in N and moisture levels of rubber leaves of various ages were determined, and their influence on the reproductive performance and the survival of L. tristis beetles was assessed. These experiments tested if the prevention of premature leaf fall in rubber plantations, a practice almost abandoned in monoculture rubber plantations due to high labor costs, is likely to enable the control of the beetle populations. It is likely to be welcomed by farmers because it is environmentally friendly and because preventing premature leaf fall would lead to higher latex production. We are unaware whether the levels of the major leaf nutrients, sodium (Na), potassium (K), calcium (Ca), and magnesium (Mg), have any role in the selection of rubber litter by the beetles. Hence, in addition to N and moisture content, Na, P, Ca, and Mg leaf levels ( Perry 1994 ; Lal et al. 2001 ) were also estimated in leaves of various ages.

Materials and Methods

Collection of beetles and the experimental set-up

n class="Species">L. tristis pupae were collected from the pan> class="Species">rubber tree plantation ( H. brasiliensis clone RRII 105) near the Devagiri college campus located at Calicut (11° 15′ N, 75° 48′ E), in the Kerala state of India in March, 2009. Teneral adults were transferred into two large circular clay vessels (13 x 35 cm) that were capped with a nylon mesh net and placed in an environmental chamber (Yorco, http://www.yorco.com ) at 70% relative humidity and 33˚ C (representing the average temperature and humidity in the rubber plantation litter). They were fed a mixture of wilted tender, senescent, and dry rubber tree leaves. To simulate the onset of summer showers starting in the last week of April, rubber leaves were no longer provided, and water was sprayed using a mist sprayer to induce dormancy. A wooden box (15 x 7 x 3 cm) was provided as the dormancy shelter for the beetles (see Sabu et al. 2008 for details).

The experiment started in the last week of December, 2009 when the beetles started showing signs of arousal from dormancy. The beetles were sexed following the sternal-notch method ( Vinod et al. 2008 ). Three leaf ages were tested: wilted tender leaves, yellow-brown senescent leaves, and dry leaves. Each replicate comprised one male-female pair in a small clay vessel (8 x 5 cm) covered with n class="Chemical">nylon mesh anpan>d kept inpan> the enpan>vironpan>menpan>tal chamber. A small, moist piece of cottonpan> placed onpan> the net served as a source of pan> class="Chemical">water, and the excreta were removed on a daily basis. The eggs produced were counted and transferred into sterile plastic vials (5.5 x 4.5 cm) using a moist fine hair brush. For each pair, the mating duration, the frequency of oviposition, and the fecundity were recorded. Ten replicates for each leaf age were maintained; thus, a total of 30 beetle pairs were analyzed. A parallel stock of 10 pairs was maintained on each leaf age to replace individuals lost by mortality during the intermediate stages of the experiment. Mortality was estimated as the number of days taken to reach 25, 50, and 100% mortality. Replacement beetles were used only for estimations of fecundity.

The 50 eggs laid during the first 24 hours of oviposition were transferred to Petri dishes (9 x 1.5 cm) in the environmental chamber and monitored at six hour intervals. The number of eggs hatched and the duration of egg development were recorded. Neonate larvae that hatched within a six hour period were transferred to labeled plastic vials (5.5 x 4.5 cm) with a moist, fine n class="Disease">hair brush. The vials were covered with finpan>e cottonpan> cloth unpan>til the emergenpan>ce of the 3 rd inpan>star larvae to prevenpan>t the escape of small larval inpan>stars anpan>d thereafter with pan> class="Chemical">nylon mesh. Ten larvae per container were maintained to follow the development of successive larval instars as well as to avoid the inhibitory effect of crowding on larval development in tenebrionids ( Tschinkel and Wilson 1971 ). Larval instars and adults were fed sliced tender rubber leaves. The durations of the larval and pupal stages were recorded.

Leaf collection

Leaves were collected from randomly selected trees from the same rubber plantation of uniform age raised from a single clone. Leaves of particular ages were not available from an individual tree for the entire course of the experiment, and hence leaves from several trees were used. By pooling leaves from several plants from the same plantation of the same clone and age, we hoped to obtain a fair estimate of nutrient concentrations. Tender leaves are distinctly different in color and size from mature leaves, which are small, brown, and smooth, and they were readily available during the early stages of foliage n class="Disease">flush. To meet the requiremenpan>t for tenpan>der leaves towards the late phase of the study, twigs of ranpan>domly selected trees were brokenpan>, anpan>d freshly sprouted leaves were collected from them. Senpan>escenpan>t yellow-brown leaves were removed by genpan>tly flickinpan>g the leaves from the trees. Freshly fallenpan> dry leaves that were brown-yellow were hanpan>dpicked directly from the upper litter layers. Senpan>escenpan>t anpan>d dry leaves were collected by trackinpan>g the trees that shed leaves late. Subsamples of the leaves collected for feedinpan>g the beetles were used for the chemical anpan>d moisture anpan>alyses.

Leaf nutrient quality estimation

Moisture content was determined by measuring the fresh weight (FW) of the leaves (to the nearest 0.001 g), drying them in paper envelopes at ambient temperature for three weeks, and re-weighing them (DW). Moisture content was calculated as (FW - DW)/FW ( Nahrung et al. 2009 ). Following the moisture analysis, the dried leaf samples from each two weeks period were pooled, oven-dried (40 o C for 3 days), ground into powder with a blender, and used for the estimations of nutrient content. n class="Chemical">Nitrogen conpan>tenpan>t was determinpan>ed followinpan>g the Kjeldahl method ( Jacksonpan> 1973 ). The estimationpan> of Na, K, Ca, anpan>d pan> class="Chemical">Mg levels was carried out following the Wet oxidation method ( Jackson 1958 ; Adler and Wilcox 1985 ) using an atomic absorption spectrophotometer (Varian AA 240 FS, Varian Medical Systems, http://www.varian.com ), and P levels were determined by the Vanado molybdate method ( Jackson 1973 ).

Statistical analysis

Ten replicates of each condition (tender, senescent, and dry leaves) were maintained. A preliminary analysis of the distribution of the data for each parameter was done with the Jarque-Bera test. The moisture content percentages were arcsine square roottransformed prior to statistical analysis. Significance levels of variation in the post-dormancy life spans of the beetles fed leaves of different ages and nutrient parameters (N, Na, K, Ca, n class="Chemical">Mg, P, anpan>d moisture conpan>tenpan>t) were anpan>alyzed with onpan>eway ANOVA tests followed by pairwise comparisonpan>s with Tukey tests. Variationpan>s inpan> fecunpan>dity durinpan>g differenpan>t phases of post-dormanpan>cy were anpan>alyzed with the Kruskal Wallis test followed by pairwise comparisonpan>s with Manpan>n-Whitney tests ( Weiss 2007 ). The inpan>fluenpan>ce of N anpan>d moisture conpan>tenpan>t onpan> biweekly fecunpan>dity of beetles fed tenpan>der leaves anpan>d onpan> the post dormanpan>cy lifespanpan> of the beetles fed leaves of various ages was examinpan>ed with multiple regressionpan> anpan>alysis. Inpan> this anpan>alysis, the qualitative variable (leaf age) was kept conpan>stanpan>t anpan>d treated as a categorical (dummy) variable. The relationpan>ship betweenpan> N anpan>d moisture conpan>tenpan>t was anpan>alyzed with the Pearsonpan> correlationpan> test to explainpan> the multicollinpan>ear relationpan>ship betweenpan> the variables. The leaf minpan>erals (peripheral variables) were excluded from the multiple regressionpan> anpan>alysis, as they lead to multicollinpan>earity amonpan>g the variables ( Graham 2003 ; Gujarati 2011 ). The signpan>ificanpan>ce levels of all anpan>alyses were p < 0.05. Minpan>itab 16 Academic software for Winpan>dows ( Minpan>itab 2010 ) was used for all statistical anpan>alyses.

Results

Beetles fed tender leaves entered into the reproductive phase, produced eggs, and survived for 135.55 ± 45.81 days, while those fed senescent and dry leaves lived for 28.25 ± 12.43 and 21.6 ± 10.36 days, respectively and did not produce eggs. The preoviposition period for beetles fed tender leaves was 13.9 ± 2.02 days, fecundity was 60.5 ± 40.23 eggs, egg laying events lasted for six months, and egg laying intervals were 6.9 ± 3.31 days.

Two phases, an initial phase of four months of fecundity and an intervening one month eggless period in the 5 th month, were distinct. The highest fecundity was recorded in the 2 nd month of the post-dormancy phase ( Figure 1 and Table 3 ). In total, 94% hatching was recorded. The duration of the egg incubation period was 3.43 ± 0.47 days, the larval instar phase lasted 33.74 ± 0.35 days, and the pupal phase lasted 3.08 ± 0.08 days .

Figure 1.

Fecundity* (mean ± SD) of post-dormancy Luprops tristis fed exclusively tender rubber leaves in relation to the phenology of Hevea brasiliensis leaf shedding (*Log transformed). The tan, brown, and light green boxes indicate falling leaves, sprouting, and light green leaves, respectively. High quality figures are available online.

Table 3.

Monthly variation in fecundity of pre-dormancy L uprops tristis in relation to the nitrogen level and moisture content of the tender leaves.

Fecundity* (mean ± SD) of post-dormancy Luprops n class="Species">tristis fed exclusively tenpan>der rubber leaves inpan> relationpan> to the phenpan>ology of pan> class="Species">Hevea brasiliensis leaf shedding (*Log transformed). The tan, brown, and light green boxes indicate falling leaves, sprouting, and light green leaves, respectively. High quality figures are available online.

Monthly variation in fecundity of pre-dormancy L uprops n class="Species">tristis inpan> relationpan> to the pan> class="Chemical">nitrogen level and moisture content of the tender leaves.

Differences in the post-dormancy life span were distinct among the three cultures (F = 108.77, DF = 2, p < 0.05). Beetles reared on dry and senescent leaves had similar life spans (p < 0.05), whereas those fed tender leaves had a significantly longer life span (p > 0.05). The times until the post-dormancy beetles reached 100% mortality were 231 days when fed tender leaves, 56 days when fed senescent leaves, and 42 days when fed dry leaves ( Table 1 and Figure 2 ).

Table 1.

Time (days) for post-dormancy Luprops tristis fed tender, senescent, and dry rubber leaves to reach 25, 50, 75, and 100% mortality.

Figure 2.

Kaplan Meier survival curve of Luprops tristis fed tender, senescent, and dry rubber leaves. High quality figures are available online.

Kaplan Meier survival curve of Luprops n class="Species">tristis fed tender, senescent, and n class="Disease">dry rubber leaves. High quality figures are available online.

Higher N content was recorded for tender leaves than senescent and dry leaves, and no variation was observed in the levels of N between senescent and dry leaves. Moisture content varied between leaf ages and was highest in tender leaves and lowest in dry leaves. Variation in the levels of the major leaf nutrients were noted, with P and K being highest in tender leaves, Ca being highest in dry leaves, and Na and n class="Chemical">Mg being highest in senescent and dry leaves ( Table 2 ). Multiple regression analysis revealed a signpan>ificant influence of the N level and moisture content on the lifespan of the post-dormancy beetles (p < 0.05; F = 55.39; R 2 = 0.89). Multiple regression analysis of the N and moisture leaf content on the fecundity of the beetles revealed a signpan>ificant influence of N ( p ≤ 0.05; T = 2.75; F = 41.73; R 2 = 0.90) and no influence of moisture ( p ≥ 0.05; T = 0.79; F = 41.73; R 2 = 0.90) on fecundity. Pearson correlation analysis revealed a high correlation between N and moisture content (r = -0.99; p ≤ 0.05).

Table 2.

Moisture content (%) and amounts of elements (mg/g) in leaves of the rubber tr ee, Hevea brasiliensis, of different ages. The different superscri pt letters within each column indicate means that differ significantly by a Student’s t -test (p < 0.05).

Moisture content (%) and amounts of elements (mg/g) in leaves of the rubber tr ee, pan> class="Species">Hevea brasiliensis, of different ages. The different superscri pt letters within each column indicate means that differ significantly by a Student’s t -test (p < 0.05).

Discussion

Reproduction and survival of post-dormancy L. tristis fed leaves of various ages

Post-dormancy n class="Species">L. tristis beetles fed senpan>escenpan>t anpan>d pan> class="Disease">dry rubber leaves exhibited significantly greater mortality and a failure to reproduce compared with those fed tender leaves, demonstrating that tender rubber leaves are essential for L. tristis to complete its life cycle. Earlier work on the link between the phenology of the rubber tree and the L. tristis life cycle revealed that the high abundance of L. tristis in rubber plantations is related to the advantages gained from feeding prematurely fallen tender leaves. Premature leaf fall is caused primarily by powdery mildew ( Odium hevea ) and Corynespora cassiicola . Hence, it was suggested that the control of the premature leaf fall in rubber plantations may enable control of the pest ( Sabu and Vinod 2009b ; Vinod and Sabu 2009; Sabu et al. 2012 ). The present study provides empirical evidence supporting these earlier predictions. Additionally, the inability of post-dormancy L. tristis to survive on senescent and dry leaves beyond 3–4 weeks suggests that reproduction depends on the availability of leaves from the premature leaf fall mediated by powdery mildew disease and not on those from the leaf fall associated with Corynespora that occurs 3–4 months later. Hence, the control of the seasonal premature leaf fall in rubber plantations caused by powdery mildew soon after leaf sprouting by spraying fungicides may enable control of this pest. These findings have great practical significance, as they reveal a strategy to tackle this pest that is otherwise not practically feasible either with beetle-directed pesticides or with natural enemies ( Aswathi and Sabu 2011 ). Furthermore, it confirms that tender leaf availability is a major limiting factor regulating the life cycle of L. tristis in the moist south Western Ghats in addition to rainfall ( Vinod and Sabu 2010 ). Since tender leaf resource availability is limited to the pre-summer period in monoculture rubber plantation belts, L. tristis would remain univoltine in the region.

The survival of post-dormancy L. tristis on dry leaves and its implications

Although post-dormancy n class="Species">L. tristis could not enpan>ter the reproductive phase, its ability to survive onpan> dry leaves for 3–4 weeks inpan>dicates its remarkable potenpan>tial to survive unpan>til leaf sproutinpan>g anpan>d subsequenpan>t tenpan>der leaf fall. This could be a strategy to counpan>teract the high mortality experienpan>ced durinpan>g the last phase of dormanpan>cy ( Sabu et al. 2008 ), which if conpan>tinpan>ued could lead to death of the enpan>tire post-dormanpan>cy beetle populationpan> anpan>d prevenpan>t the productionpan> of the next genpan>erationpan>. Onpan>e quarter of the beetles perish durinpan>g the 9 monpan>th dormanpan>cy period (see Sabu et al. 2008 for details). The presenpan>t record of 75% mortality of post-dormanpan>cy pan> class="Species">L. tristis fed dry leaves within 3-4 weeks indicates that only one-quarter of post-dormancy L. tristis returning to rubber plantations could survive and enter the reproductive phase upon the return of tender leaves. The high survival rate of post-dormancy L. tristis fed tender leaves suggests that the availability of tender leaves of other rubber clones in the RI 115 plantation (other host plants are unlikely in monoculture rubber plantations) would lead to higher survival rates for the post-dormancy L. tristis and a rise in population. These findings provide an answer to the questions raised in earlier studies ( Sabu et al. 2007 ; Sabu and Vinod 2009a ) on whether the post-dormancy L. tristis that return to the plantations could survive on dry leaves during the initial phase of leaf fall in rubber plantations. They indicate perfect synchronization of the life cycle of the beetle with host plant phenology at two occasions— first at the time of post-dormancy return and the annual leaf shedding by rubber trees and later at the time of entry into the breeding phase and the premature leaf fall in rubber plantations ( Sabu and Vinod 2009a ).

Leaf age-related variations in mortality indicate that if annual leaf shedding and tender leaf availability are delayed, one could expect high mortality of post-dormancy n class="Species">L. tristis . Conpan>versely, if the anpan>nual leaf sheddinpan>g starts prematurely due to the early cessationpan> of the monpan>soonpan>s anpan>d the onpan>set of summer conpan>ditionpan>s, there would be low mortality anpan>d larger populationpan>s. Such variationpan>s inpan> the anpan>nual leaf sheddinpan>g anpan>d tenpan>der leaf availability could have caused the variationpan> inpan> the abunpan>danpan>ce of the pan> class="Species">L. tristis population and the intensity of infestation during certain years. Monitoring the variations in leaf shedding would enable the predictions of the severity of infestation necessary to initiate precautionary measures to limit the intensity of home invasions.

The high fecundity and prolonged post-dormancy phase of L. tristis fed tender leaves

The fecundity of beetles fed exclusively tender leaves in laboratory conditions was higher than the fecundity in natural conditions ( Sabu et al. 2008 ). Hence, upon premature leaf fall, the duration of the post-dormancy phase and fecundity will increase (60.5 ± 40.24 eggs in this study under ideal conditions, in contrast to 30.6 ± 13.92 eggs under natural conditions; Sabu et al. 2008 ) leading to larger populations and severe n class="Disease">beetle aggregation. However, these conpan>ditionpan>s lead to the emergenpan>ce of tenpan>eral adults with less time for food reserve accumulationpan> anpan>d low survival chanpan>ces durinpan>g dormanpan>cy ( Sabu et al. 2008 ). Currenpan>tly, the RR 115 rubber clonpan>e with early leaf fall anpan>d leaf sproutinpan>g durinpan>g Janpan>uary is beinpan>g replaced by the RR 414, RR 424, anpan>d RR 430 rubber clonpan>es, which display delayed leaf sheddinpan>g anpan>d leaf sproutinpan>g durinpan>g February. The combinpan>ed effect of the late leaf sproutinpan>g of the new clonpan>es anpan>d the early leaf sheddinpan>g of the old RR 115 clonpan>es will lead to a prolonpan>ged period of tenpan>der leaf availability unpan>til the complete replacemenpan>t occurs over a 10–15 year period. Henpan>ce, a further rise inpan> the pan> class="Species">L. tristis population in this region is predicted.

A decline in fecundity towards the late phase of post-dormancy even when tender leaves are available indicates that fecundity variation during the post-dormancy n class="Species">L. tristis phase canpan>not be attributed to the leaf quality variationpan>. Inpan>stead, this result could reflect anpan> adaptive mechanpan>ism of pan> class="Species">L. tristis to synchronize egg production and the feeding phase with tender leaf resource availability to produce a new generation of beetles (class 1 type described earlier; see Sabu et al. 2008 for details) with more food reserves and better survival chances during the forthcoming dormancy phase. The eggless period towards the last phase of post-dormancy corresponds to the period of home invasion and the onset of rainfall. What leads to cessation of egg laying during this period is not understood, and it could be linked to the inherent genetic disposition towards dormancy (Denlinger 1986; Leather et al. 1995 ).

Leaf age-related variations in rubber leaf quality

Variations in the levels of major nutrients and moisture content occur in deciduous n class="Species">rubber tree leaves of various ages, anpan>d the highest levels occur inpan> tenpan>der leaves, likely due to the leaf nutrienpan>t resorptionpan> mechanpan>ism of senpan>escinpan>g leaves ( Killinpan>gbeck 1996 ; Ecksteinpan> et al. 1998 ; Aerts anpan>d Chapinpan> 2000 ; Vanpan> Heerwaardenpan> et al. 2003 ). The nutrienpan>t resorptionpan> mechanpan>ism is conpan>sidered onpan>e of the most importanpan>t planpan>t nutrienpan>t conpan>servationpan> mechanpan>isms ( Wright anpan>d Westoby 2003 ; Yanpan> et al. 2006 ; Huanpan>g et al. 2007 ). Foliar nutrienpan>t conpan>cenpan>trationpan>s remainpan> relatively conpan>stanpan>t from the time of full leaf expanpan>sionpan> to the beginpan>ninpan>g of senpan>escenpan>ce anpan>d thenpan> decrease rapidly as foliar nutrienpan>ts are resorbed prior to abscissionpan> ( Hevia et al. 1999 ). N, P, anpan>d K are mobile nutrienpan>ts that are easily withdrawn from senpan>escinpan>g tissues, anpan>d K is known for leachinpan>g ( Perry 1994 ; Lal et al. 2001 ; Hagenpan>-Thornpan> et al. 2006 ). Henpan>ce, the fall inpan> the levels of N anpan>d P inpan> older leaves is attributed to inpan>tenpan>sive nutritive resorptionpan>, anpan>d the fall inpan> K levels is likely due to leachinpan>g loss durinpan>g the prolonpan>ged monpan>soonpan> period inpan> additionpan> to resorptionpan>. Earlier studies suggest that pan> class="Chemical">Mg is moderately resorbed (up to 20%), whereas Ca is not resorbed prior to leaf abscission ( Hagen-Thorn et al. 2006 ). Ca is an immobile nutrient, leading to its higher concentration ( Epstein 1972 ; Perry 1994 ; Lal et al. 2001 ) in senescent leaves. These data indicate that rubber is a ‘‘nutrient conservative’’ species with high nutritive resorption during leaf senescence, and the senescent and dry leaves are therefore of lower nutrient quality compared to tender leaves.

The effects of leaf age-related variations in chemical quality on the post-dormancy survival and fecundity of L. tristis

The present study reveals that tender leaves are essential for n class="Species">L. tristis to enpan>ter inpan>to the reproductive phase anpan>d complete its life cycle anpan>d that the levels of N anpan>d pan> class="Chemical">water, the two most important nutritional components for growth, are high in tender leaves; furthermore, herbivore performance (survival, growth, and reproductive capacity) was also high in tender leaves ( Mattson 1980 ; Raupp and Denno 1983 ; Osier and Lindroth 2001 ; Awmack and Leather 2002; Holton et al. 2003 ). The high correlation between N and moisture content indicates that due to the multicollinearity between the variables in the regression analysis, the N values are masking the effect of moisture content on fecundity; otherwise, the moisture content would have been a significant contributor. No broad generalizations about the impacts of other minerals on the longevity and fecundity of L. tristis can be made from these data, as other nutrients and defensive components of the diet modify their effects ( Martel 1998 ; Clancy 1992 ).

Conclusions

The results of the experiments, though conducted in controlled conditions, could be used to forecast the performance of n class="Species">L. tristis inpan> field conpan>ditionpan>s anpan>d could be inpan>cluded inpan> decisionpan>s onpan> populationpan> manpan>agemenpan>t anpan>d conpan>trol. The presenpan>t study shows that the availability of tenpan>der leaves with high levels of N anpan>d moisture is a critical factor that determinpan>es the fate of adult post-dormanpan>cy pan> class="Species">L. tristis beetles and the survival of the next generation. Since post-dormancy L. tristis obtains tender leaves from the powdery mildew-mediated premature leaf fall, the control of the premature leaf fall will reduce the population of the next generation of beetles. Additionally, these results imply that because tender leaves are essential for the reproductive maturity of L. tristis and are available only for a limited period of time, because wet conditions drive the beetles indoors, it is highly likely that upon removing tender leaves, L. tristis will remain univoltine in the rubber plantation belts.