Biological Control 53 (2010) Contents lists available at ScienceDirect. Biological Control. journal homepage:

Transcription

1 Biological Control 53 (2010) Contents lists available at ScienceDirect Biological Control journal homepage: Evaluating the impact of herbivory by a grasshopper, Cornops aquaticum (Orthoptera: Acrididae), on the competitive performance and biomass accumulation of water hyacinth, Eichhornia crassipes (Pontederiaceae) Angela Bownes a,b, *, Martin P. Hill b, Marcus J. Byrne c a Agricultural Research Council-Plant Protection Research Institute (ARC-PPRI), Private Bag X6006, Hilton 3245, South Africa b Department of Zoology and Entomology, P.O. Box 94, Rhodes University, Grahamstown 6140, South Africa c Department of Animal, Plant and Environmental Sciences, University of the Witwatersrand, Private Bag X3, Wits 2050, Johannesburg, South Africa article info abstract Article history: Received 19 August 2009 Accepted 26 February 2010 Available online 4 March 2010 Keywords: Agent efficacy Plant competition Inverse linear model The water hyacinth grasshopper, Cornops aquaticum is being considered for release in South Africa as a biocontrol agent for water hyacinth, Eichhornia crassipes. The release decision was to be based on C. aquaticum s potential efficacy, in addition to its host specificity. An additive series analysis of competition between E. crassipes and a similar free-floating macrophyte, water lettuce (Pistia stratiotes), was used to assess the impact of C. aquaticum on E. crassipes competitive ability and potential for biomass accumulation. The data was analyzed using an inverse linear model where the competitive performance of both plant species was estimated using multiple linear regressions of the inverse of biomass yield. Eichhornia crassipes was 24 times more competitive than P. stratiotes without C. aquaticum herbivory but this was reduced to 12 in the presence of herbivory which equated to a 50% reduction in its competitive performance. Pistia stratiotes competitive ability increased from 3.8 to 4.8 (i.e. 21%) when E. crassipes was damaged by C. aquaticum. The interspecific competition coefficients from E. crassipes on P. stratiotes were also no longer statistically significant in the presence of C. aquaticum herbivory. Biomass accumulation of E. crassipes was significantly reduced by C. aquaticum herbivory at both planting densities tested. By comparison with similar studies on two of the most abundant, widespread and damaging biocontrol agents already released in South Africa, C. aquaticum has a greater impact on E. crassipes, suggesting it will be a valuable addition to the South African biological control program. Ó 2010 Elsevier Inc. All rights reserved. 1. Introduction The biological control program on water hyacinth, Eichhornia crassipes Mart. Solms-Laubach (Pontederiaceae), in South Africa has been active for almost 40 years. Since its initiation, five arthropods and one pathogen have been released in an attempt to reduce infestations to manageable levels (Hill and Cilliers, 1999). Despite these efforts, together with a substantial investment of resources into controlling E. crassipes with other methods, it remains South Africa s most economically, socially and environmentally problematic aquatic weed (Van Wyk and van Wilgen, 2002). For this reason, additional agents are under consideration, with the water hyacinth grasshopper, Cornops aquaticum Brüner (Orthoptera: Acrididae) having been prioritized. * Corresponding author. Address: Agricultural Research Council-Plant Protection Research Institute (ARC-PPRI), Private Bag X6006, Hilton 3245, South Africa. Fax: +27 (0) address: BownesA@arc.agric.za (A. Bownes). Cornops aquaticum is native to South America and was first introduced into quarantine in South Africa in 1995 for host specificity testing which showed it to be oligophagous, utilizing species in the family Pontederiaceae (Oberholzer and Hill, 2001). A release permit was granted in However, it was also considered important to evaluate C. aquaticum in terms of its potential to suppress E. crassipes populations. The science of biological control has undergone changes in approach and practice over the last decade and pre-release efficacy testing is now considered a critical aspect of pre-release studies (Balciunas, 2004; McClay and Balciunas, 2005). In the history of weed biocontrol, many agents have become established and built up good numbers but have failed to have an impact on the target plant (Myers, 2000; McClay and Balciunas, 2005). As a result, biocontrol practitioners are increasingly under pressure to improve predictions on success and to determine the potential impact of a candidate agent on the target plant prior to release (Pantone et al., 1989; Sheppard, 2003; Balciunas, 2004; McClay and Balciunas, 2005). The ultimate aim of this approach is to reduce the number of organisms being purposefully introduced into new environments while at the same time, increase /$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi: /j.biocontrol

2 298 A. Bownes et al. / Biological Control 53 (2010) success rates of biocontrol programs (Pantone et al., 1989; McEvoy and Coombs, 1999; Sheppard, 2003). Two processes that influence the abundance and distribution of plant species in a community are competition and herbivory (Harper, 1977) and they are therefore potentially important processes in the control of invasive alien plants. Crawley (1983) notes that the primary effect of herbivory on plants is not to eat them to extinction, but rather that it modifies the competitive ability of one plant species with another. Biocontrol practitioners are increasingly realizing the potential benefits of combining these two processes and that an interaction between the two can amplify the negative effects of herbivory by a biocontrol agent. Plant performance might only be reduced by herbivory in the presence of interspecific competition (Cottam et al., 1986; Rand, 2004) and herbivory can mediate exploitative plant competition through selective herbivory on the dominant competitor (Cottam, 1986). Furthermore, success in some biocontrol programs has been attributed to a combination of herbivory and a change in the competitive status of the plant (Whittaker, 1979), and biocontrol studies have shown that interactions between competition and herbivory are often synergistic, having a greater impact on the target plant than either factor working in isolation (Cottam et al., 1986; Ang et al., 1994; Van et al., 1998). The relative importance of plant competition and herbivory in structuring plants communities will depend on prevailing conditions such as nutrient availability to plants (Maschinski and Whitham, 1989; Swank and Oechel, 1991) or plant biomass (Bonser and Reader, 1995) through influencing competitive ability or potential for compensation. Pantone et al. (1989) developed a method for evaluating efficacy of biocontrol agents where the change in competitive interactions between the target plant and a similar plant species is measured in the presence and absence of selective herbivory by the biocontrol agent. Inverse linear models are used to analyze data from additive series competition experiments (Spitters, 1983). Center et al. (2001) suggested the method for evaluating biocontrol agents as it provides experimental data that can be used for comparing the impact of one agent with another. McClay and Balciunas (2005) have subsequently suggested it as a good method for evaluating efficacy of candidate biological control agents. Therefore, as part of a holistic approach in determining potential efficacy of C. aquaticum, an adaptation of Pantone et al. s (1989) model for E. crassipes, where water lettuce, Pistia stratiotes L. (Araceae) is used as a competitor (Center et al., 2001), was used to assess the effect of herbivory on E. crassipes competitive ability. More importantly, as noted by Coetzee et al. (2005), utilization of the method provides biologically meaningful competition coefficients that enable direct comparison with other E. crassipes biocontrol agents evaluated in similar competition studies. Efficacy of three E. crassipes biocontrol agents widely established in South Africa, the sap-sucking mirid, Eccritotarsis catarinensis Caravalho (Hempitera: Miridae) (Coetzee et al., 2005) and the water hyacinth weevils, Neochetina bruchi Hustache (Coleoptera: Curculionidae) and N. eichhornia Warner (Coleoptera: Curculionidae) as well as a combination of the two (Ajuonu et al., 2008) have already been assessed using this method. Comparison of the damage potential of C. aquaticum compared with the Neochetina weevils and E. catarinensis was considered crucial for the release decision in order to determine whether C. aquaticum has greater potential to control E. crassipes, thereby warranting its release. The aim of the paper was therefore to: (1) evaluate the impact of C. aquaticum herbivory on the competitive performance and biomass accumulation of E. crassipes; (2) use an inverse linear model to make direct comparisons with the results from other studies which evaluated three biocontrol agents already established in South Africa and (3) discuss the significance of the findings in terms of whether or not to release C. aquaticum. 2. Materials and methods The experimental design was based on that used by Coetzee et al. (2005). Eichhornia crassipes and P. stratiotes plants obtained from stock cultures were grown in plastic tubs in a polycarbon glasshouse at the Plant Protection Research Institute in Pretoria, South Africa. The tubs (63 cm 42 cm 37 cm) were filled with 20 L of water. To test the response of E. crassipes to C. aquaticum herbivory when the plant s growth rates are optimal (Reddy et al., 1989), conditions which resembled a highly eutrophic impoundment were created by adding nitrates and phosphates to the water as potassium nitrate (KNO 3 ) and potassium dihydrogen orthophosphate (KH 2 PO 4 ), respectively at a rate of 7.6 mg N L 1 and 1.37 mg P L 1 (Ryan Brudvig, pers. comm.). Chelated iron was added at a rate of 1.7 g/20 L water (Coetzee et al., 2005). The nutrient medium and water were replaced weekly and each tub was enclosed with a net canopy. The canopies were constructed of a fine screening mesh (aperture size = 1 mm) and were designed to confine the grasshoppers to the tubs. The trial ran from November to December The design followed an additive series (Spitters, 1983) of factorial combinations of different densities of the two competing plant species. The planting densities of E. crassipes:p. stratiotes were 0:3, 0:9, 3:0, 3:3, 3:9, 9:0, 9:3 and 9:9 in each tub. The density matrices were repeated twice, once as the experiment with herbivory and once as the control without herbivory and were replicated three times. The resulting 48 tubs were arranged in a randomized block design (Coetzee et al., 2005). Plants were grown for 2 weeks prior to the start of the trial, after which, all daughter plants were removed to revert to the initial stocking densities. Adult C. aquaticum were added at one pair per two E. crassipes plants, which is the density at which they occur under dense field populations in the region of origin (Silveira-Guido and Perkins, 1975). Because of an uneven number of plants in every tub, the extra grasshopper added was a female. Insect densities were maintained at the original density by removing emerging nymphs as soon after hatching as possible. After 4 weeks, the plants in each tub were weighed to determine total wet weight (including daughter plants and dead plant material). These values for each plant species in each tub were then divided by the original plant stocking density to calculate mean wet weight per individual E. crassipes and P. stratiotes plant. The inverse linear model described by Spitters (1983), and Pantone et al. (1989) was used for the analysis. The competitive ability of each species was estimated using multiple linear regressions of the inverse of the mean weight-yield of each species as the dependent variable and the planting densities of each species as the independent variables. The regression equation is of the form: 1=Wh ¼ a h0 þ a hh d h þ a h1 d 1 1=Wl ¼ a l0 þ a ll d 1 þ a lh þ a lh d h (Center et al., 2001), where 1/Wh is the inverse biomass yield of individual E. crassipes plants and 1/Wl is the inverse biomass yield of individual P. stratiotes plants. The respective planting densities are represented by d h and d 1. Intraspecific competition is estimated by the coefficients, a hh and a ll, and interspecific competition is estimated by a hl and a lh in terms of their effects on the reciprocals of the yield of both plant species. The intercepts are represented by a h0 and a l0 which are the reciprocals of the maximum weight of isolated plants. The ratio of the coefficients a hh /a h1 measures the effects of intraspecific competition of E. crassipes on itself, relative to the effects of interspecific competition of P. stratiotes on E. crassipes. Likewise, the ratio of the coefficients a ll /a lh measure the effects of intraspecific competition by P. stratiotes on its own yield relative

3 A. Bownes et al. / Biological Control 53 (2010) to the effect of interspecific competition by E. crassipes on the yield of P. stratiotes (Pantone et al., 1989). The graphical representations of the data with surface response planes show the combined effect of P. stratiotes and E. crassipes planting densities on the reciprocal yield of both plant species. Data of the mean end-weights (total yield/original planting density) of E. crassipes and P. stratiotes were normally distributed. A three-way analysis of variance (ANOVA) was used to investigate the effects of E. crassipes and P. stratiotes planting densities (competition treatment), herbivory by C. aquaticum and their interactions on biomass yield of both E. crassipes and P. stratiotes. 3. Results 3.1. Impact of herbivory on competition Eichhornia crassipes Eichhornia crassipes was the superior competitor of the two plant species in the presence and absence of herbivory by C. aquaticum (Table 1). In the absence of herbivory, E. crassipes was 24 times as strong a competitor with itself as was P. stratiotes with E. crassipes, indicating that intraspecific competition has a much greater effect on E. crassipes than interspecific competition from P. stratiotes. In the herbivory treatment, the ratio of the competition coefficients (a hh /a hl ) was only 12 which equates to a 50% reduction in E. crassipes competitive ability due to selective herbivory by C. aquaticum. Interspecific competition coefficients were not significant in the presence or absence of C. aquaticum herbivory (Table 1). However, competition from P. stratiotes did increase in the presence of C. aquaticum herbivory, as indicated by the coefficient a hl (Table 1), which is a fourfold increase in P. stratiotes competitive ability, although the total effect on E. crassipes biomass is still small. The graphical representation of the data provides a visual indication of this increase through the slightly steeper slope of the P. stratiotes regression (Fig. 1A) in the herbivory treatment compared to the relatively flat slope in the control (Fig. 1B). This indicates that without herbivore pressure, P. stratiotes had no effect on E. crassipes yield. Intraspecific competition had a much greater impact on E. crassipes yield (Table 1). The steeper slopes of the regression planes on the E. crassipes density axes (Fig. 1A and B) for both the herbivory treatment and the control are a visual indication that competition from conspecifics had a much greater impact on E. crassipes biomass than competition from P. stratiotes Pistia stratiotes Pistia stratiotes was a much weaker competitor compared to E. crassipes. Without herbivory by C. aquaticum, adding 3.8 (a ll / a lh = 3.752) P. stratiotes plants had the same effect on P. stratiotes biomass as adding one E. crassipes plant. However, in the presence of selective herbivory on E. crassipes, the ratio of the competition coefficients increased to 4.8 indicating a 21% increase in the competitive ability of P. stratiotes, when E. crassipes plants were stressed by herbivory. More importantly, the interspecific competition coefficients evaluating the effect of E. crassipes on P. stratiotes were no longer significant in the presence of selective herbivory by C. aquaticum on E. crassipes (Table 1). The intraspecific competition coefficients (a ll ) are almost identical (Table 1) which is expected, considering that P. stratiotes plants were not subjected to damage from herbivory. It is evident from the graphical representation of the data that the planting densities of both species had a significant impact on P. stratiotes yield indicated by the steep slopes in both directions (Fig. 2). There was only a slight decrease in the interspecific competition coefficient (a lh )(Table 1) in the herbivory treatment compared to the control, as is visible from the surface response planes (Fig. 2A and B). These show a marginally steeper slope in P. stratiotes yield without herbivory compared to the herbivory regression. Both intra- and interspecific competition were acting on P. stratiotes biomass Impact of herbivory and competition on plant biomass (mean endweights) Intraspecific competition and herbivory by C. aquaticum had a highly significant effect on E. crassipes biomass yield (Table 2). High levels of competition at an E. crassipes planting density of nine, and the presence of herbivory significantly reduced the mean endweights per original E. crassipes plant (Fig. 3A). Interspecific competition from P. stratiotes had no effect on E. crassipes yield and none of the interactions between competition and herbivory were significant (Table 2). The mean end-weights per original P. stratiotes plant were significantly reduced by intraspecific competition (Fig. 3B and Table 2) and interspecific competition from E. crassipes (Table 2). The C. aquaticum herbivory treatment had no effect on P. stratiotes biomass yield and none of the interactions between competition and herbivory were statistically significant (Table 2). 4. Discussion Eichhornia crassipesis clearly the dominant species compared with P. stratiotes and it remained the superior competitor in both the presence and absence of herbivory by C. aquaticum in these trials. Agami and Reddy (1990) demonstrated the superior competitive ability of E. crassipes compared to P. stratiotes using a Table 1 Multiple regression analysis of the effects of herbivory by Cornops aquaticum and plant density on the reciprocal of Eichhornia crassipes yield and Pistia stratiotes yield (wet weight (kg); significant competition coefficients in bold). Treatment Regression coefficients Intercept R 2 F-value a hh (t value; P) a hl (t value; P) a hh /a hl a h0 E. crassipes (1/Wh) Herbivory (13.60; <0.0001) (1.14; 0.268) (P < ) Control (11.68; <0.0001) ( 0.49; 0.627) (P < ) P. stratiotes (1/Wl) Herbivory (9.94; <0.0001) (2.07; ) (P < ) Control (15.46; <0.0001) (4.11; ) (P < ) 1/Wh: The intercept a h0 estimates the reciprocal of the maximum weight of isolated E. crassipes plants. The regression coefficients a hh and a h1 measure intra- and interspecific competition, respectively for E. crassipes. The ratio a hh /a h1 measures the effects of intraspecific competition by E. crassipes on its own weight relative to the effects of interspecific competition by P. stratiotes. 1/Wl: The intercept a l0 estimates the reciprocal of the maximum weight of isolated P. stratiotes plants. The regression coefficients a ll and a lh measure intra- and interspecific competition, respectively for P. stratiotes. The ratio a ll /a lh measures the effects on intraspecific competition by P. stratiotes on its own weight relative to the effects of interspecific competition by E. crassipes.

4 300 A. Bownes et al. / Biological Control 53 (2010) Fig. 1. Multiple regression planes indicating the effects of Cornops aquaticum herbivory and plant densities on the inverse of the mean wet weight (kg) per Eichhornia crassipes plant. (A) and (B)compare relative competitive abilities of E. crassipes in the presence and absence of C. aquaticum herbivory, respectively. Points indicate observations (n=18) and the vertical lines between the data points indicate the residuals. Values on the X and Y axes are the original planting densities of E. crassipes and Pistia stratiotes. reciprocal replacement series of E. crassipes and P. stratiotes planting densities. The results reported here are consistent with similar studies evaluating the impact of E. crassipes biocontrol agents on the competitive performance of E. crassipes, when in competition with P. stratiotes (Center et al., 2005; Coetzee et al., 2005; Ajuonu et al., 2008). Under favorable nutrient conditions, E. crassipes luxuriant growth and high plasticity (Agami and Reddy, 1990) enable it to flourish even in the presence of another weed species. Plants Fig. 2. Multiple regression planes indicating the effects of Cornops. aquaticum herbivory and plant densities on the inverse of the mean wet weight (kg) per Pistia stratiotes plant. (A) and (B) compare relative competitive abilities of P. stratiotes in the presence and absence of Cornops aquaticum herbivory, respectively. Points indicate observations (n=18) and the vertical lines between the data points indicate the residuals. Values on the X and Y axes are the original planting densities of Eichhornia crassipes and P. stratiotes. with the advantage of phenotypic plasticity, such as E. crassipes, can respond to changes in the distribution of resources and, through phenotypic adjustment, for example in leaf area, extension of petioles or allocations to root biomass, can maximize capture of resources such as light and nutrients (Grime, 1979). Under high nutrient conditions, E. crassipes shoot:root ratio is 2:1, compared to only 1:1 in P. stratiotes, which demonstrates the superior ability of the plant for acquisition of resources such as light (Agami and Reddy, 1990). These characteristics contribute to the competitive

5 A. Bownes et al. / Biological Control 53 (2010) Table 2 Summary of analyses of variance (ANOVA) investigating the effects of Eichhornia crassipes (E) and Pistia stratiotes (P) planting densities, Cornops aquaticum herbivory (C) and their interactions on E. crassipes and P. stratiotes biomass yield (significant values in bold). Source MS F-value P-value E. crassipes EC density (E) < PS density (P) C. aquaticum (C) < H L H C L C H L C P. stratiotes EC density (E) < PS density (P) < C. aquaticum (C) H L H C L C H L C Fig. 3. Mean end-weights (kg) of individual Eichhornia crassipes (A) and Pistia stratiotes (B) plants in the herbivory treatments and the controls at planting densities of 3 and 9. Means followed by the same letter are not significantly different. Error bars represent the standard error of the mean. dominance of E. crassipes over similar species and its notoriety as a highly aggressive competitor in the aquatic environment (Wright and Purcell, 1995). Despite maintaining dominance in the presence of selective herbivory, there was a significant decrease in E. crassipes competitive ability after only 4 weeks of herbivory by C. aquaticum. Intraspecific competition in E. crassipes was reduced by 50% and the effects of interspecific competition of E. crassipes on P. stratiotes were significantly reduced when E. crassipes was damaged by C. aquaticum herbivory. The decline in E. crassipes competitive performance as a result of grasshopper herbivory indicates that C. aquaticum has a significant impact on E. crassipes vigor. The combined biomass data also indicated that defoliation by C. aquaticum significantly reduced the accumulation of biomass at both planting densities compared to control plants within the 4-week period. A reduction in E. crassipes biomass and intraspecific competitive ability suggests that C. aquaticum has the potential to reduce the density and spread of E. crassipes infestations in South Africa, should it be released. One of the principle aims of using this methodology to assess the potential efficacy of C. aquaticum was to have quantitative values that could be compared with other studies that have performed equivalent measures on other biocontrol agents of E. crassipes. In conforming to new strategies in biological control of weeds, where new agents are released because they are necessary and have the potential to be superior biocontrol agents (McEvoy and Coombs, 1999; Pearson and Callaway, 2005), it was important to assess C. aquaticum compared to the other biocontrol agents. While three of these biocontrol agents evaluated using the inverse linear model caused significant reductions in E. crassipes competitive ability, this was measured over longer periods and at densities at which they occur under field conditions in areas of introduction, as opposed to the native-range density of C. aquaticum tested here. Coetzee et al. (2005) found a 56% reduction in E. crassipes competitive ability after 16 weeks of herbivory by the sap-sucking mirid, E. catarinensis at a density of 15 per plant, but no differences in plant biomass as a result of herbivory. However, 40 mirids per plant reduced E. crassipes competitive ability by 101% and caused reductions in biomass after only 8 weeks (Ajuonu et al., 2008). Herbivory by the two water hyacinth weevils, N. eichhorniae and N. bruchi decreased E. crassipes competitive performance by 98% (Center et al., 2005) after 10 weeks at a density of four weevils per plant. A combination of N. eichhorniae and E. catarinensis at half the density that was used in single species trials had the greatest impact on E. crassipes competitive ability, reducing it by 229% after 8 weeks of herbivory suggesting a synergism between the two agents (Ajuonu et al., 2008). The most significant finding from the C. aquaticum study is a 50% reduction in E. crassipes competitive ability and significant reductions in plant biomass after only 4 weeks of herbivory and at a density of only one grasshopper per plant. This study maintained a constant insect density whereas the preceding examples allowed test populations to increase by reproduction. Neochetina bruchi was found to be more damaging than N. eichhorniae which was partly attributed to its higher fecundity and shorter development time, therefore having faster rates of population growth than its congener (Center et al., 2005). If the emerging C. aquaticum nymphs had been allowed to remain on plants and the study period extended, the plants would not, from personal observation, have survived much longer at elevated insect densities. The nymphs are extremely damaging and at the nutrient levels used in this trial, high numbers of nymphs, as a result of high fecundity of females (Bownes, 2009), would most likely have contributed to rapid defoliation of E. crassipes plants within a few days of emergence of the nymphs. Therefore the grasshopper s influence on competitive ability at conservative densities and without a population increase indicates that C. aquaticum has a great potential to reduce E. crassipes competitive dominance.

6 302 A. Bownes et al. / Biological Control 53 (2010) Herbivory by C. aquaticum and intraspecific competition were the dominant factors influencing E. crassipes biomass. Intraspecific competition in weed species can be intense, for example high densities of knapweed reduced individual plant biomass and shoot number (Müller-Schärer, 1991). Although the negative effects of herbivory on E. crassipes were independent of interspecific competition, the greatest reductions in E. crassipes biomass were evident at a high E. crassipes density in combination with high competitor densities and herbivory. These results are also in accordance with the biomass-dependent theory of Grime (1979) where increasing levels of both intra- and interspecific competition had a negative effect on E. crassipes biomass and are consistent with the findings of Bonser and Reader (1995) and Tiffin (2002) that negative competition effects are increased with increasing plant density. Resources become more limiting and plants suffer reductions in fitness associated with resource-limitation. The effect of the grasshopper in significantly reducing E. crassipes biomass at both planting densities tested indicates that the plant is not able to effectively compensate for C. aquaticum herbivory in the presence of high or low levels of intraspecific competition. Furthermore, there are two factors that suggest that the effects of C. aquaticum herbivory on E. crassipes observed in the present study might be greater in the field in South Africa. Firstly, compensatory ability of E. crassipes for C. aquaticum herbivory in terms of its potential to replace lost tissue is higher under eutrophic nutrient conditions (Bownes, 2009). The nitrate level used in this trial is similar to Reddy et al. s (1989) nitrate level which they found to be the threshold concentration for maximum E. crassipes growth and productivity. Water nutrient data collected over a 2-year period at 15 E. crassipes sites around South Africa indicated that all sites with the exception of Mbozambo Swamp had average nitrate levels below 2.5 mg L 1 (Byrne et al., in press) which is on the border of eutrophic and mesotrophic. Under these nutrient conditions, E. crassipes growth and productivity is significantly reduced (Reddy et al., 1989; Coetzee et al., 2007; Bownes, 2009), therefore, a greater impact on E. crassipes biomass and competitive ability could be expected in most E. crassipes sites around the country where C. aquaticum establishes. Second is the fact that these reductions in E. crassipes performance and vigor were at a conservative density of one grasshopper per plant. In the absence of C. aquaticum s specialist predator in its region of origin, Ludovix fasciatus Gyllenhal (Coleoptera: Curculionidae) (Silveira-Guido and Perkins, 1975), South Africa should expect higher population densities of C. aquaticum than those found in its native range. An evaluation of density damage relationships between C. aquaticum and E. crassipes showed that the ability of E. crassipes to compensate for C. aquaticum herbivory under eutrophic nutrient conditions decreases linearly with an increase in C. aquaticum densities (Bownes, 2009). Therefore all aspects related to the invasiveness of E. crassipes such as its rapid growth rates and competitive dominance, are likely to be reduced in the presence of population densities greater than one grasshopper per plant. One of the most important findings from this research is that the negative impact that C. aquaticum herbivory has on E. crassipes does not depend on the presence of a competing plant species and that the grasshopper has the potential to reduce E. crassipes vigor and biomass accumulation in a very short period of time. It also enabled comparison with data from similar studies that evaluated E. crassipes biocontrol agents already released in South Africa. C. aquaticum is only eligible for release if it has proven potential to be a superior biocontrol agent. Based on this comparison and its impact on E. crassipes, C. aquaticum has potential to contribute to control of E. crassipes in South Africa. These results will however, be collated with other experimental data on potential efficacy, and C. aquaticum E. crassipes interactions as part of a holistic approach in assessing whether C. aquaticum should be introduced into the South African biological control program for E. crassipes. Acknowledgements The Working for Water (WfW) Programme of the Department of Water Affairs and Forestry, South Africa is thanked for providing funding to conduct this research. We also thank Lawrence Maphoso for assistance with setting up the experiment. References Agami, M., Reddy, K.R., Competition for space between Eichhornia crassipes (Mart.) Solms and Pistia stratiotes L. cultured in nutrient-enriched water. Aquatic Botany 38, Ajuonu, O., Byrne, M.J., Hill, M.P., Neuenschwander, P., Korie, S., The effect of two biological control agents, the weevil Neochetina eichhorniae and the mirid Eccritotarsus catarinensis on E. Crassipes, Eichhornia crassipes, grown in culture with water lettuce, Pistia stratiotes. Biocontrol 54, Ang, B.N., Kok, L.T., Holtzman, G.I., Wolf, D.D., Competitive growth of Canada thistle, tall fescue and crownvetch in the presence of a thistle defoliator, Cassida rubiginosa Müller (Coleoptera: Chrysomelidae). Biological Control 4, Balciunas, J., Are mono-specific agents necessarily safe? The need for prerelease assessment of the probably impact of candidate biological control agents, with some examples. In: Cullen, J.M., Briese, D.T., Kriticos, D.J., Lonsdale, W.M., Morin, L., Scott, J.K. (Eds.), Proceedings of the XI International Symposium on the Biological Control of Weeds. CSIRO Publishing, Melbourne, Australia, pp Bonser, S.P., Reader, R.J., Plant competition and herbivory in relation to vegetation biomass. Ecology 76, Bownes, A., Evaluation of a plant herbivore system in determining potential efficacy of a candidate biological control agent, Cornops aquaticum for E. crassipes, Eichhornia crassipes. PhD Thesis. Rhodes University, Grahamstown. Byrne, M.J., Hill, M.P., Robertson, M., King, A, Jadhav, A., Katembo, M., Wilson, J., Brudvig, R., Fisher, J., in press. Integrated management of E. crassipes in South Africa: development of an integrated management plan for E. crassipes control, combining biological control, herbicidal control and nutrient control, tailored to the climatic regions of South Africa. Water Research Commission, Pretoria, South Africa. Center, T.D., Van, T.K., Hill, M.P., Can competition experiments be used to evaluate the potential efficacy of new E. crassipes biocontrol agents. In: Julien, M.P., Hill, M.P., Center, T.D., Jianqing, D. (Eds.), Proceedings of the Second Meeting of the Global Working Group for the Biological and Integrated Control of E. crassipes. ACIAR, Australia, pp Center, T.D., Van, T.K., Dray Jr., F.A., Franks, S.J., Rebelo, M.T., Pratt, P.D., Rayamajhi, M.B., Herbivory alters competitive interactions between two invasive aquatic plants. Biological Control 33, Coetzee, J.A., Center, T.D., Byrne, M.J., Hill, M.P., Impact of the biocontrol agent Eccritotarsus catarinensis on the biological control of E. Crassipes, Eichhornia crassipes. Aquatic Botany 86, Coetzee, J.A., Byrne, M.J., Hill, M.P., Impact of nutrients and herbivory by Eccritotarsus catarinensis on the biological control of E. Crassipes, Eichhornia crassipes. Aquatic Botany 86, Cottam, D.A., The effects of slug-grazing on Trifolium repens and Dactylis glomerata in monoculture and mixed sward. Oikos 47, Cottam, D.A., Whittaker, J.B., Malloch, A.J.C., The effects of chrysomelid grazing and plant competition on the growth of Rumex obtusifolius. Oecologia 70, Crawley, M.J., Herbivory. The Dynamics of Animal Plant Interactions. Blackwell Scientific Publications, Oxford. Grime, J.P., Plant Strategies and Vegetative Processes. John Wiley and Sons, Chichester. Harper, J.L., Population Biology of Plants. Academic Press, New York. Hill, M.P., Cilliers, C.J., A review of the arthropod natural enemies, and factors that influence their efficacy, in the biological control of E. Crassipes, Eichhornia crassipes (Mart.) Solms-Laubach (Pontederiaceae), in South Africa. African Entomology Memoir 1, Maschinski, J., Whitham, T.G., The continuum of plant responses to herbivory: the influence of plant association, nutrient availability and timing. The American Naturalist 134, McClay, A.S., Balciunas, J.K., The role of pre-release efficacy assessment in selecting classical biological control agents for weeds applying the Anna Karenina principle. Biological Control 35, McEvoy, P., Coombs, Biological control of plant invaders: regional patterns, field experiments and structured population models. Ecological Applications 9, Müller-Schärer, H., The impact or root herbivory as a function of plant density and competition, survival, growth and fecundity of Centaurea maculosa in field plots. Journal of Applied Ecology 28,

7 A. Bownes et al. / Biological Control 53 (2010) Myers, J.H., What can we learn from biological control failures? In: Spencer, N.R. (Ed.), Proceedings of the X International Symposium on Biological Control of Weeds. USDA-ARS, Montana, pp Oberholzer, I.G., Hill, M.P., How safe is the grasshopper Cornops aquaticum for release on E. crassipes in South Africa. In: Julien, M.H., Hill, M.P., Center, T.D., Jianqing, D. (Eds.), Proceedings of the Second Meeting of the Global Working Group for the Biological Control and Integrated Control of E. crassipes. ACIAR, Australia, pp Pantone, D.J., Williams, W.A., Maggenti, A.R., An alternative approach for evaluating the efficacy of potential biocontrol agents of weeds. 1. Inverse linear model. Weed Science 37, Pearson, D.E., Callaway, R.M., Indirect nontarget effects of host-specific biological control agents: implications for biological control. Biological Control 35, Rand, T.A., Competition, facilitation and compensation for insect herbivory in an annual salt marsh forb. Ecology 85, Reddy, K.R., Agami, M., Tucker, J.C., Influence of nitrogen supply rates on growth and nutrient storage by E. Crassipes (Eichhornia crassipes) plants. Aquatic Botany 36, Sheppard, A.W., Prioritizing agents based on predicted efficacy. In: Jacob, H.S., Briese, D.T. (Eds.), Improving the Selection. Testing and Evaluation of Weed Biological Control Agents. CRC for Australia Weed Management, Glen Osmond, pp Silveira-Guido, A., Perkins, D.B., Biology and host specificity of Cornops aquaticum (Bruner) (Orthoptera: Acrididae), a potential biological control agent for E. Crassipes. Environmental Entomology 4, Spitters, C.J.T., An alternative approach to the analysis of mixed cropping experiments. 1. Estimation of competition effects. Netherlands Journal of Agricultural Science 31, Swank, S.E., Oechel, W.C., Interactions among the effects of herbivory, competition and resource limitation on chaparral herbs. Ecology 72, Tiffin, P., Competition and timing of damage affect the pattern of selection acting on plant defenses against herbivores. Ecology 83, Van, T.K., Wheeler, G.S., Center, T.D., Competitive interactions between hydrilla (Hydrilla verticillata) and Vallisneria (Vallisneria americana) as influenced by insect herbivory. Biological Control 11, Van Wyk, E., van Wilgen, B.W., The cost of E. Crassipes control in South Africa: a case study of three options. African Journal of Aquatic Science 27, Whittaker, J.B., Invertebrate grazing, competition and plant dynamics. In: Anderson, R.M., Turner, B.D., Taylor, L.R. (Eds.), Population Dynamics. The 20th Symposium of the British Ecological Society, London, Blackwell Scientific Publications, Oxford. Wright, A.D., Purcell, M.F., Eichhornia crassipes (Mart.) Solms-Laubach. In: Groves, R.H., Shepherd, R.C.H., Richardson, R.G. (Eds.), The Biology of Australian Weeds. R.G. and F.G Richardson, Melbourne, pp