Ecol Res (2006) 21:617–623 DOI 10.1007/s11284-006-0159-9 O R I GI N A L A R T IC L E Marcelo N. Rossi Æ Carolina Reigada Wesley A.C. Godoy The effect of hunger level on predation dynamics in the spider Nesticodes rufipes: a functional response study Received: 23 August 2005 / Accepted: 10 January 2006 / Published online: 20 May 2006 ! The Ecological Society of Japan 2006 Abstract It is well known that a predator has the potential to regulate a prey population only if the predator responds to increases in prey density and inflicts greater mortality rates. Predators may cause such densitydependent mortality depending on the nature of the functional and numerical responses. As spiders are usually faced with a shortage of prey, the killing behavior of the spider Nesticodes rufipes at varying densities of Musca domestica was examined here through laboratory functional response experiments where spiders were deprived of food for 5 (well-fed) or 20 days (hungry). An additional laboratory experiment was also carried out to assess handling time of spiders. The number of prey killed by spiders over 24- and 168-h periods of predator–prey interaction was recorded. Logistic regression analyses revealed the type II functional response for both well-fed and hungry spiders. We found that the lower predation of hungry spiders during the first hours of experimentation was offset later by an increase in predation (explained by estimated handling times), resulting in similarity of functional response curves for well-fed and hungry spiders. It was also observed that the higher number of prey killed by well-fed spiders over a 24-h period of spider–prey interaction probably occurred due to their greater weights than hungry spiders. We concluded that hungry spiders may be more voracious than well-fed spiders only over longer time periods, since hungry spiders may spend more time handling their first prey items than well-fed spiders. M. N. Rossi (&) Departamento de Botânica, IB, Universidade Estadual Paulista (Unesp), Botucatu, Sao Paulo, Brazil E-mail: mnrossi@ibb.unesp.br Tel.: +55-1438116265 Fax: +55-1438153744 C. Reigada Æ W. A. Godoy Departamento de Parasitologia, IB, Universidade Estadual Paulista (Unesp), Botucatu, Sao Paulo, Brazil Keywords Nesticodes rufipes Æ Musca domestica Æ Functional response Æ Predator–prey interaction Æ Poultry house Æ Shortage of prey Introduction To evaluate the role of spiders in regulating insect populations in many natural systems, it is vital to know their foraging response when faced with fluctuations in prey population densities (Riechert and Lockley 1984; Wise 1993). In the case of increasing prey populations, spiders can react in two ways: either by increasing the consumption by individuals (functional response) or by increasing their own density (numerical response). These concepts were proposed by Solomon (1949) and developed further by Holling (1959). The functional response is defined as the temporal rate at which an individual predator kills prey. Thus, the functional response is a double rate: it is the average number of prey killed per individual predator per unit of time (Holling 1959; Turchin 2003). Most studies of this response have been performed in the laboratory, since field studies are technically difficult (Marc et al. 1999). In type I functional responses, the number of prey killed increases linearly to a maximum then remains constant as prey density increases. This corresponds to a constant proportion of available prey killed to the maximum number (density independence), followed by a declining proportion of prey killed. In type II functional responses, the number of prey killed approaches the asymptote hyperbolically as prey density increases. This corresponds to an asymptotically declining proportion of prey killed (inverse density dependence). In type III functional responses, the number of prey killed approaches the asymptote as a sigmoid function. This corresponds to an increase in the proportion of prey killed (density dependence) to the inflection point of the sigmoid curve, followed by a decrease in the proportion of prey killed (Juliano 2001). 618 The type II functional response is predominant for spiders, whereas types I and III are less common (Heong et al. 1991; Mansour and Heimbach 1993; Samu and Biro 1993; Wise 1993; Denno et al. 2004). However, the type of functional response exhibited by spiders is quite variable. For example, Riechert and Lockley (1984) observed the presence of type III functional response for several spiders, even though of the five papers that they cite, three provide no clear support for a type III response (Kiritani and Kakiya 1975; Mansour et al. 1980). Nakamura (1977) found both type II and III responses by lycosids, and the most pronounced type III response occurred with leafhopper prey. The author argued that accelerated rates of predation at higher densities resulted from greater prey activity due to increased interference among leafhoppers in the rice seedlings in the experimental container. Haynes and Sisojevic (1966) also observed type III functional response for a crab spider due to increased activity of the prey at higher densities. Shortage of prey seems to be a very important selective factor in the evolution of foraging strategies of spiders because spiders can survive for long periods without eating (Framenau et al. 2000). Most researches that have studied the effects of a varying food supply on foraging spiders concentrate on energetic investment, such as changes in foraging site, or adjustments to the architecture of the web (Janetos 1982; Provencher and Riechert 1991; Bradley 1993; Lubin and Henschel 1996; Herberstein et al. 2000). However, only a few studies have investigated the effect of different hunger levels on the killing behavior of spiders (Holling et al. 1980; Bridge and Wootton 1998; Framenau et al. 2000; Herberstein et al. 1998, 2000). Nesticodes rufipes (Lucas) (Araneae: Theridiidae) (referred to as Theridion rufipes in references) is widely distributed in tropical and subtropical regions, extending to temperate zones; the spiders construct irregular webs with a disordered aspect (González 1989). Their exact distribution is not easy to determine, since they are strongly associated with humans (Downes 1988; González and Estévez 1988; González 1989). Behavioral and ecological studies considering predation by N. rufipes are scarce. Fox (1998) highlighted the strategic importance of these spiders in the natural control of the mosquito Aedes aegypti (L.) (Diptera: Culicidae), since the spiders incorporate a paralyzing substance in their webs that paralyzes the mosquitoes on contact, increasing their capture efficiency. Barreto et al. (1987) also mentioned the importance of N. rufipes as predators of Rhodnius prolixus (Stal) (Hemiptera: Reduviidae). Musca domestica (L.) (Diptera: Muscidae) has a cosmopolitan distribution and high synanthropic indices (Smith 1986; Ferreira and Lacerda 1993), and also is of considerable medical and veterinary importance (Harwood and James 1979; Smith 1986; Levine and Levine 1991). This species lives in human dwellings, poultry houses, supermarkets and garbage, being reared on a wide variety of substrates such as food and vertebrate excrement (Axtell and Arends 1990; Bowman 1995). Although there are some chemical techniques aimed at controlling M. domestica in poultry houses, inappropriate application of chemicals has intensified the search for potential natural enemies of houseflies (Cunha and Lomônaco 1996). Therefore, understanding the strength of interspecific interactions between M. domestica and its predators is of major importance. As M. domestica (adults) is usually seen in N. rufipes webs in poultry houses (Rossi and Godoy 2005), it is necessary to characterize the type of functional response exhibited by this predator when houseflies are offered as prey in order to determine the inherent ability of this predator to regulate densities of prey population. Yet, since spiders are usually faced with prey shortage, it is important to understand the underlying mechanisms of the killing behavior that spiders experience at different hunger levels. Thus, the specific objectives of this study were to (1) investigate the ability of ‘‘well-fed’’ and ‘‘hungry’’ N. rufipes to kill housefly individuals (voracity) and assess characteristics of their predation dynamics (functional response) during the course of predator–prey interaction, and (2) estimate handling time exhibited by well-fed and hungry spiders as a basis for understanding the killing behavior observed. Materials and methods Sampling and rearing of houseflies and spiders An experimental poultry house located in the city of Botucatu, SP (Brazil) (22!52¢20¢¢S; 48!26¢37¢¢W) was chosen to collect larvae of houseflies. We removed larvae from small samples of chicken feces deposited below the cages, and put them into small glass tubes. All insects were then taken to the laboratory and reared in vials containing wet ground animal ration (25!C under 12-h light). After pupation, vials were kept in cages of nylon mesh on a wood frame (30·30·30 cm) where water and sugar were provided for adults. Adult females of N. rufipes were captured in several buildings located on the campus of the University of the State of São Paulo (Botucatu, Brazil) from January– March 2003, and kept individually in cylindrical plastic containers (5.0 cm in height · 8.5 cm of diameter) in the laboratory (25!C under 12-h light). All spiders were of similar size (15 mm of body length) and were fed with adult houseflies for 1 month (insects were randomly offered twice a week) in order to attain similar nutritional status. Functional response study For the functional response experiment, spiders were placed individually in large cylindrical plastic containers [11.5 cm in height · 10.5 cm of diameter (900 ml)], with one spider per container, with a nylon mesh (10·3 cm) internally fixed in each container in order to 619 allow spiders to build their webs. As spiders usually fixed their small webs right below the lids during the period of experimentation, the containers were suitable for carrying out the experiments. We observed that larger containers would reduce predator–prey interaction significantly as only a small fraction of prey available for consumption would be captured. The experiment was a randomized complete block (5·2 factorial) with five prey densities (3, 5, 10, 15 or 20 adult flies per container) and two hunger levels of spiders (deprived of food for 5 and 20 days). Each treatment combination was repeated 15 times with different spiders for each replicate. After the respective periods without eating, spiders were individually weighed (grams) by using a semianalytical scale, and a Student’s t-test (Zar 1999) was run comparing the mean weights of spiders deprived of food for 5 (well-fed) and 20 days (hungry). Immediately before adding prey into the spider containers, flies were immobilized by chilling in a freezer for 3 min, removed from cages, and placed in a Petri dish. When flies started to move, all insects were carefully dropped in the bottom of a spider container, without touching the spider web. This procedure prevented flies from being captured quickly due to their flying ability and it insured that flies could be easily separated prior to the experiments. In the first 2 min (approximately) inside the spider containers, flies just walked and then started flying. Spider and prey interacted for 168 h (7 days), after which we determined the number of prey killed (sucked and wrapped insects) by subtracting the number of remaining insects (living and intact dead individuals) from the density of prey offered. To reduce the mortality of flies during the experiment, water and sugar were provided ad libitum for adults (piece of cotton soaked in honeydew solution). Dead flies fixed on the web that were not sucked or not even wrapped were also computed as intact dead individuals. In order to keep densities of flies constant, intact dead individuals were replaced with new living flies in each container at 24-h intervals during the 168 h of experimentation. As spiders generally kill more prey than they consume, we defined functional response in terms of the number of prey individuals killed, rather than eaten. We also recorded the number of prey killed over 24 h of predator–prey interaction because it was expected that all spiders would be well-fed after consumption of the first prey items, even those spiders deprived of food for 20 days. Therefore, we predicted that the handling time and consumption of prey by spiders from different hunger levels could vary during experimentation, changing the predation dynamic. The time of 168 h for predator–prey interaction was necessary because it provided enough data (number of prey killed) for fitting to the functional response model. To assess possible variation in the predation dynamic during the course of spider–prey interaction, three ANOVA [5·2 factorial design (see above)] were run. The first two ANOVA compared the mean number of prey killed over 24 and 168 h of spider–prey interaction, respectively (Zar 1999). The third ANOVA also compared the mean number of prey killed by spiders over 168 h of interaction; however, data from the first 24 h of interaction were not considered in this case. Therefore, we could compare the voracity (the ability of spiders to kill prey) between well-fed and hungry spiders at distinct moments. When significant differences were found between hunger levels, Student’s t-tests (Zar 1999) were run to compare mean values. We determined the relationship between prey density and the number of prey killed (functional response) by spiders at both hunger levels by performing logistic regressions (PROC CATMOD) on the proportion of prey killed vs. density of prey offered (Trexler et al. 1988; Juliano 2001; SAS 2001). For predator functional response, logistic regression is particularly useful in distinguishing between type II and type III responses, which are not easily determined by nonlinear regressions that use the number of prey eaten as the dependent variable. The proportion of prey eaten declines monotonically with prey density in a type II response but is positively density-dependent over some range of prey densities in a type III response. The sign of the linear coefficient estimated by the logistic regression can be used to distinguish the shape of the functional response curve. Significant negative or positive linear coefficients from the regression indicate type II or type III, respectively (Trexler et al. 1988; Messina and Hanks 1998; De Clercq et al. 2000; Juliano 2001). Because logistic regression analysis indicated that our data fit the type II response for spiders in both hunger levels, further analysis was restricted to the type II response. The ‘‘random predator equation’’ (Royama 1971; Rogers 1972) was used to describe the functional responses because it allows for prey depletion during the course of the experiment. The form of the type II equation is as follows: Ne ¼ N f1 " exp ½"aðT " Th Ne Þ&g where Ne is the number of prey killed; N, the number of prey offered; T, the total time available for the predator, which in this case is 168 h; a is the attack rate; and Th is the handling time. Using the SAS statistical package (SAS 2001) we carried out nonlinear regressions, using the least squares method to estimate the predator’s attack rate (a) and handling time (Th) considering predation by spiders at both hunger levels. Further, plots showing the observed and predicted number (data fitted to random predator equation) of prey killed at a given prey density were also determined for both hunger levels. Laboratory assessment of handling time To estimate the handling time (Th) directly, an additional experiment was set up. As above, spiders were removed from the plastic rearing containers, and placed individually in large plastic containers [11.5·10.5 cm (900 ml)]. Five prey densities (3, 5, 10, 15 or 20 adult flies per container) were offered to spiders exposed to two hunger levels (deprived of food for 5 and 20 days). Each treatment combination was repeated five times with different spiders for each replicate. Containers were filmed for 24 h in the laboratory (25!C under 12-h light), and images recorded on a Time Lapse videocassette. We estimated handling time directly only during 24 h of experimentation because it was the maximum time that our cameras could shoot prey consumption accurately. As nine-color cameras were available for shooting, a multiplexer was coupled to the video. Therefore, the monitor (29¢¢) was divided in nine quadrates with each receiving images from a specific container. Each camera was rested on an iron support positioned at a standard distance of 5 cm from the top of the container. Thus, images were obtained with accurate resolution. Immediately before shooting, prey was added to the containers as described above. As all containers could not be shot simultaneously, each treatment combination was selected randomly. Time from the capture of prey until its abandonment by the predator was considered handling time. Correlations of handling time with prey densities were run for well-fed and hungry spiders. Finally, the nonparametric Mann-Whitney U-test (Zar 1999) was computed, comparing the handling time between well-fed and hungry spiders. Results Comparison of mean weights of spiders showed that spiders deprived of food for 20 days weighed less than spiders deprived of food for 5 days (t=5.992, df=148, P<0.05, n=75 for each hunger level). During 24 h of predator–prey interaction, the mean number of prey killed by spiders differed between hunger levels (F=18.140, MS=41.294, P<0.0001), as hungry spiders killed significantly less prey than did well-fed spiders (Fig. 1a). Even though the mean number of prey killed among densities was also significant (F=5.550, MS=12.633, P<0.0005), data did not fit well into the random predator model (r2<0.20 for both hunger levels). Therefore, only data of prey killed over 168 h of predator–prey interaction were fitted to the model. During 168 h of predator–prey interaction, differences in prey killed between hunger levels were not significant when all data were considered (F=2.220, MS=11.162, P=0.1386); however, after removing data from the first 24 h of interaction, the mean number of prey killed was significantly different (F=22.861, MS=95.394, P<0.0001). Curiously, hungry spiders killed significantly more prey than well-fed spiders in this case (Fig. 1b). The mean number of prey killed among densities was significant for both situations (F=62.467, MS=314.104, P<0.0001; and F=49.851, MS=202.020, P<0.0001). Mean (+/-SD) number of prey killed 620 10 a) 8 Well-fed spiders Hungry spiders b) 6 4 2 0 Hunger level Fig. 1 Mean (±SD) number of adults of Musca domestica killed by spiders deprived of food for 5 (well-fed) and 20 (hungry) days. a Comparison computed over 24 h of predator–prey interaction. Means differed statistically according to Student’s t-test (t=4.178, df=142, P<0.05). b Comparison computed over 168 h of predator–prey interaction without considering data from the first 24 h of interaction. Again, means differed statistically by Student’s t-test (t=!2.770, df=142, P<0.05). For each hunger level, n=72 (three spiders died during experimentation) Logistic regression analyses indicated a type II functional response for both hunger level conditions (Fig. 2), and estimates of the linear coefficients were significantly different from zero (P<0.01; Table 1). Model-estimated attack rates for well-fed and hungry spiders were 0.20 and 0.40 h!1, respectively. Modelestimated handling times for well-fed and hungry spiders were 18.22 and 14.09 h, respectively. Correlations of direct estimation of handling times with initial prey numbers offered were also not significant for well-fed (r2=0.1608, P=0.4426, n=25), and hungry spiders (r2=!0.0548, P=0.7949, n=25). Handling times estimated directly were significantly different between well-fed [3.68 h (median); 2.46 (quartile range)] and hungry spiders [5.08 h (median); 3.32 (quartile range)], and hungry spiders spent significantly more time handling a prey item (U=209.0, P<0.05, n=25 for each hunger level). Discussion Nesticodes rufipes exhibited a type II functional response at both hunger levels against M. domestica. However, parameters estimated by the random predator equation demonstrated that the well-fed spiders had a lower attack rate and greater handling time than the hungry spiders over 168 h of predator–prey interaction. Consequently, differences in attack rate and handling time indicated that hungry spiders were more voracious while capturing and killing prey. Nevertheless, the number of prey killed by spiders at both hunger levels was not significantly different over 168 h of interaction. Although the same type of functional response was observed for both hunger levels, predation data for hungry spiders presented a greater coefficient of determination (Fig. 2b) than that of well-fed spiders 621 Number of prey killed 16 (a) 14 r2 = 0.46 12 10 8 6 4 2 0 0 4 8 12 16 Number of prey offered 20 24 20 24 Number of prey killed 16 (b) 14 r2 = 0.53 12 10 8 6 4 2 0 4 8 12 16 Number of prey offered Fig. 2 Functional response of Nesticodes rufipes to increasing densities of Musca domestica during 168-h period. Logistic regressions showed that spiders at both hunger levels exhibited a type II functional response across the range of prey densities offered. a Spiders deprived of food for 5 days (well-fed). b Spiders deprived of food for 20 days (hungry). Each data point represents the observed number of M. domestica killed. Curves were fitted using the random predator equation (see Materials and methods) (Fig. 2a), indicating a higher variation of predation for the latter. It may explain why many studies of functional response set up in the laboratory keep predators under deprivation of food for some pre-established periods, since in this case, data may fit better to models. Hence, it is very important to know how often specific predators go without feeding in nature; otherwise there is a risk of such experiments showing inaccurate data. We observed that N. rufipes usually fixed their small webs right below the lids of the containers during the period of experimentation. With webs covering a limited space, this particular container size enabled spiders to catch good numbers of prey, therefore, data fitted properly to the functional response model. As we were interested in testing the effect of the hunger level on predation dynamics of this spider, this container size was suitable to investigate this question. According to Fox (1998) this spider incorporates a paralyzing substance in the webs, which increases prey capture. However, we did not observe such an effect in our study. In fact, most flies escaped from webs when not captured or bitten quickly by spiders. It is possible that such a paralyzing substance may have detectable effects only on small or lightweight prey. Physiological responses of spiders to starvation suggest they have experienced food shortages frequently throughout their evolutionary history. Older instars of numerous species can survive long periods without feeding (Anderson 1974; Wise 1993). Spiders survive when deprived of prey by maintaining a relatively motionless sit-and-wait foraging strategy, and by decreasing their metabolic rate (Nakamura 1972; Anderson 1974; Tanaka and Itô 1982; Tanaka et al. 1985). Thus, in our study 20 days was the longest time that spiders went without eating, which is much closer to natural situations. Although our laboratory-measured functional response of N. rufipes may not exactly correspond to field situations, this study presents important results because females of N. rufipes may experience long periods without consuming any prey in the field. It is known that a predator has the potential to regulate densities of its prey only if the mortality rate it inflicts is density-dependent, which can occur if the predator displays a type III functional response (Holling 1959; Wise 1993). However, robust type III functional responses are probably not common among spiders. Many results suggest that most type III responses by spiders result from elevated prey activity at higher prey densities rather than from learning or modification in foraging behavior by spiders (Haynes and Sisojevic 1966; Nakamura 1977; Riechert and Lockley 1984). Our data corroborate most studies of functional response by spiders (Kajak 1978; Heong et al. 1991; Mansour and Heimbach 1993; Samu and Biro 1993; Wise 1993; Finke and Denno 2002) because we observed the type II functional response, at least for the densities of prey offered, which implies lack of regulation of M. domestica population by N. rufipes. Table 1 Results of logistic regression analysis of the proportion of adults of Musca domestica killed by Nesticodes rufipes at two hunger levels over 168 h of predator–prey interaction against initial prey numbers offered (all analyses were significant at P<0.01) Hunger level Parameter Estimate SE v2 P Deprived of food for 5 days Intercept Linear Quadratic Cubic Intercept Linear Quadratic Cubic 9.2236 !1.7733 0.1350 !0.00354 14.1757 !2.9586 0.2285 !0.00584 2.4801 0.6208 0.0489 0.00121 4.1679 0.9715 0.0725 0.00173 13.83 8.16 7.61 8.54 11.57 9.27 9.93 11.42 0.0002 0.0043 0.0058 0.0035 0.0007 0.0023 0.0016 0.0007 Deprived of food for 20 days 622 As already discussed, the mean number of prey killed did not differ between well-fed and hungry spiders over 168 h of predator–prey interaction (Fig. 2). However, over 24 h of interaction, the risk of predation of M. domestica adults by N. rufipes differed markedly between spiders at both hunger levels. Well-fed spiders killed significantly more prey than hungry spiders (Fig. 1a), and this is explained by their significantly lower handling times. Perhaps, as hungry spiders spent more time on each captured prey, they ingested more food from a lower number of prey than well-fed spiders, resulting in different tactics of food ingestion between the two spider groups. Basically, data from this study indicated that the killing behavior of N. rufipes toward M. domestica is dynamic and may be dependent on the hunger level, because during the first 24 h of interaction, well-fed spiders spent less time handling prey, and consequently killed more prey, while hungry spiders, according to the model, spent less time handling prey during 168 h of interaction. Therefore, if we had observed predation only during the first 24 h of predator–prey interaction, it would lead us to conclude that well-fed spiders were more voracious than hungry spiders. However, lower predation of hungry spiders during the first hours of experimentation was offset later by an increase in predation, justifying similar functional response curves (Fig. 2a, b). In conclusion, we believe that the higher number of prey killed by well-fed spiders during 24 h of predator– prey interaction occurred due to their greater weights than those of hungry spiders, which resulted in more storage of resources and increasing fitness for prey capture. However, after eating the first prey items, hungry spiders also became well-fed and then killed more prey than well-fed spiders due to having previously experienced shortage of prey. This is a very interesting finding, since it implies that hungry spiders may be more voracious than well-fed spiders only over longer time periods, since hungry spiders may spend more time handling their first prey items than well-fed spiders. Acknowledgements We thank Professors José Ricardo and James Welsh for providing statistical advice and for revising the text of the manuscript, respectively. M.N. Rossi and C. Reigada are particularly grateful to Fapesp (Fundação de Amparo à Pesquisa do Estado de São Paulo) for financial support. W.A.C. Godoy has been supported by a research fellowship from Conselho Nacional de Desenvolvimento Cientı́fico e Tecnológico. References Anderson JF (1974) Responses to starvation in the spiders Lycosa lenta Hentz and Filistata hibernalis (Hentz). Ecology 55:576– 585 Axtell RC, Arends JJ (1990) Ecology and management of arthropod pests of poultry. Annu Rev Entomol 35:101–126 Barreto M, Barreto P, D’Alessandro A (1987) Predation on Rhodnius prolixus (Hemiptera: Reduviidae) by the spider Theridion rufipes (Araneida: Theridiidae). J Med Entomol 24:115– 116 Bowman DD (1995) Parasitology for veterinarians. WB Saunders, Philadelphia Bradley RA (1993) The influence of prey availability and habitat on activity patterns and abundance of Argiope keyserlingi (Araneae: Araneidae). J Arachnol 21:91–106 Bridge AP, Wootton RW (1998) Effects of satiation and food deprivation on feeding behaviour in the pholcid spider, Pholcus phalangioides. Bull Br Arachnol Soc 11:107–110 Cunha CL, Lomônaco C (1996) Monitorização de inpacto ambiental provocado por dispersão de moscas em bairros adjacentes a uma granja avı́cola. An Soc Entomol Brasil 25:1–12 De Clercq P, Mohaghegh J, Tirry L (2000) Effect of host plant on the functional response of the predator Podisus nigrispinus (Heteroptera: Pentatomidae). Biol Control 18:65–70 Denno RF, Mitter MS, Langellotto GA, Gratton C, Finke DL (2004) Interactions between a hunting spider and a web-builder: consequences of intraguild predation and cannibalism for prey suppression. Ecol Entomol 29:566–577 Downes MF (1988) The effect of temperature on oviposition interval and early development in Theridion rufipes (Araneae: Theridiidae). J Arachnol 16:41–45 Ferreira MJM, Lacerda PV (1993) Muscóides sinantrópicos associados ao lixo urbano em Goiânia, GO. Rev Bras Zool 10:185– 195 Finke DL, Denno RF (2002) Intraguild predation diminished in complex-structured vegetation: implications for prey suppression. Ecology 83:643–652 Fox I (1998) Predation on Aedes aegypti (Diptera: Culicidae) by Theridion rufipes (Araneae: Theridiidae) in Puerto Rico. J Med Entomol 35:611–613 Framenau VW, Finley LA, Allan K, Love M, Shirley D, Elgar MA (2000) Multiple feeding in wolf spiders: the effect of starvation on handling time, ingestion rate, and intercatch intervals in Lycosa lapidosa (Araneae: Lycosidae). Aust J Zool 48:59–65 González A (1989) Análisis del comportamiento sexual y producción de ootecas de Theridion rufipes (Araneaea: Theridiidae). J Arachnol 17:129–136 González A, Estévez AL (1988) Estudio del desarrollo postembrionario y estadı́sticos vitales de Theridion rufipes Lucas, 1846 (Araneae: Theridiidae). Rev Bras Entomol 32:499–506 Harwood RF, James MT (1979) Entomology in human and animal health. Macmillan, New York Haynes DL, Sisojevic P (1966) Predatory behaviour of Philodromus rufus Walckenaer (Araneae: Thomisidae). Can Entomol 98:113–133 Heong KL, Bleih S, Rubia EG (1991) Prey preference of the wolf spider, Pardosa pseudoannulata (Boesenberg et Strand). Res Popul Ecol 33:179–186 Herberstein ME, Abernethy KE, Backhouse K, Bradford H, deCrespigny FE, Luckock PR, Elgar MA (1998) The effect of feeding history on prey capture behaviour in the orbweb spider Argiope keyserlingi Karsch (Araneae: Araneidae). Ethology 104:565–571 Herberstein ME, Craig C, Elgar MA (2000) Foraging strategies and feeding regimes: web decoration investment in Argiope keyserlingi Karsch (Araneae: Araneidae). Evol Ecol Res 2:69– 80 Holling CS (1959) The components of predation as revealed by a study of small-mammal predation of the European sawfly. Can Entomol 91:293–320 Holling CS, Hardman JM, Turnbull AL (1980) Functional response of the wolf spider, Pardosa vancouveri, to changes in the density of vestigial-winged fruit flies. Res Popul Ecol 21:233– 259 Janetos AC (1982) Active foragers vs. sit-and-wait predators: a simple model. J Theor Biol 95:381–385 Juliano SA (2001) Nonlinear curve fitting: predation and functional response curves. In: Scheiner SM, Gurevitch J (eds) Design and analysis of ecological experiments. Oxford University Press, New York, pp 178–196 Kajak A (1978) Analysis of consumption by spiders under laboratory and field conditions. Ekol Pol 26:409–427 623 Kiritani K, Kakiya N (1975) An analysis of the predator–prey system in the paddy field. Res Popul Ecol 17:29–38 Levine OS, Levine MM (1991) Houseflies (Musca domestica) as mechanical vectors of shigellosis. Rev Infect Dis 13:688–696 Lubin Y, Henschel J (1996) The influence of food supply on foraging behaviour in a desert spider. Oecologia 105:64–73 Mansour F, Heimbach U (1993) Evaluation of lycosid, micryphantid and linyphiid spiders as predators of Rhopalosiphum padi (Hom.: Aphididae) and their functional response to prey density—laboratory experiments. Entomophaga 38:79–87 Mansour F, Rosen D, Shulov A (1980) Functional response of the spider Chiracanthium mildei (Arachnida: Clubionidae) to prey density. Entomophaga 25:313–316 Marc P, Canard A, Ysnel F (1999) Spiders (Araneae) useful for pest limitation and bioindication. Agric Ecosyst Environ 74:229–273 Messina FJ, Hanks JB (1998) Host plant alters the shape of the functional response of an aphid predator (Coleoptera: Coccinelidae). Environ Entomol 27:1196–1202 Nakamura K (1972) The ingestion in wolf spiders. II. The expression of degree of hunger and amount of ingestion in relation to spider’s hunger. Res Popul Ecol 14:82–96 Nakamura K (1977) A model for the functional response of a predator to varying prey densities based on the feeding ecology of wolf spiders. Bull Natl Inst Agric Sci Jpn Ser C 31:28–89 Provencher L, Riechert SE (1991) Short-term effects of hunger conditioning on spider behavior, predation, and gain of weight. Oikos 62:160–166 Riechert SE, Lockley T (1984) Spiders as biological control agents. Annu Rev Entomol 29:299–320 Rogers D (1972) Random search and insect population models. J Anim Ecol 41:369–383 Rossi MN, Godoy WAC (2005) Web contents of Nesticodes rufipes and Latrodectus geometricus (Araneae: Theridiidae) in a Brazilian poultry house. J Entomol Sci 40:347–351 Royama T (1971) A comparative study of models for predation and parasitism. Res Popul Ecol (Suppl 1):1–91 Samu F, Biro Z (1993) Functional response, multiple feeding and wasteful killing in a wolf spider (Araneae: Lycosidae). Eur J Entomol 90:471–476 SAS (2001) SAS/STAT software; version 8.2. SAS Institute, Cary Smith KGV (1986) A manual of forensic entomology. University Printing House, Oxford Solomon ME (1949) The natural control of animal populations. J Anim Ecol 18:1–35 Tanaka K, Itô Y (1982) Decrease in respiratory rate in a wolf spider, Pardosa astrigera (L. Koch), under starvation. Res Popul Ecol 24:360–374 Tanaka K, Itô Y, Saito T (1985) Reduced respiratory quotient by starvation in a wolf spider, Pardosa astrigera (L. Koch). Comp Biochem Physiol 80A:415–418 Trexler JC, McCulloch CE, Travis J (1988) How can the functional response best be determined? Oecologia 76:206–214 Turchin P (2003) Complex population dynamics: a theoretical empirical synthesis. Monographs in population biology, vol. 35. Princeton University Press, Princeton Wise DH (1993) Spiders in ecological webs. Cambridge University Press, Cambridge Zar JH (1999) Biostatistical analysis. Prentice-Hall, Upper Saddle River