Pea landraces may hold genetic variation that can be exploitable in breeding/selecting new cultivars. In a small-seeded pea landrace, four phenotypes were distinguished according to testa (green, non-spotted and green, spotted) and hilum colour (white, black). The four phenotypes were tested for two growing seasons in the field for pod (seeds/pod) and seed traits (1000-seed weight, toughness, total phenols and tannins, testa colour, protein and carbon concentration, C/N ratio, bruchid infection). Significant differences were found for testa colour parameters, phenolic, tannin and carbon concentration and bruchid tolerance. The larger-seeded, spotted peas had darker testa and more phenols, while white-hilumed peas had lighter testa and more tannins. The spotted, black-hilumed phenotype, with the highest carbon concentration and C/N ratio was the most tolerant to bruchids. However, grouping the phenotypes, neither spotted nor black-hilumed ones showed to be more tolerant compared with their counterparts. Concluding, our results showed that phenotyping variation in seeds of a pea landrace revealed variation in seed traits, which could be exploitable. Since testa and hilum colour were associated with specific seed traits, they could, alone or in combination, be used as biomarkers of seed quality traits in pea. Testing of larger number of phenotypes is needed to solidify our findings.

It has been suggested that soil treated with a herbicide and subsequently carried by wind and deposited on plant foliage can cause crop injury. This study compared foliar uptake and translocation of herbicides applied to plants as an aqueous solution or in herbicide-treated soil. Leaves of 3-wk-old seedling alfalfa, grape, and pea were treated with 14C-labeled thifensulfuron, chlorsulfuron, glyphosate, 2,4-D, and bromoxynil. Significant amounts of all herbicides were absorbed by pea, alfalfa, and grape from the aqueous solutions, whereas very limited absorption occurred from herbicide-treated soil. Prolonged and multiple exposure to herbicide-treated soil did not increase herbicide uptake. High relative humidity enhanced herbicide absorption from aqueous solutions but not from herbicide-treated soil. All herbicides except bromoxynil were readily translocated in alfalfa, grape, and pea. Limited quantities of herbicides were absorbed from herbicide-treated soil by plant foliage, and this small amount is unlikely to cause crop damage.

The effects of atrazine [2-chloro-4-(ethylamino)-6-(isopropylamino)-s-triazine], cyanazine {2-[[4-chloro-6-(ethylamino)-s-triazin-2-yl] amino]-2-methylpropionitrile}, and procyazine {2-[[4-chloro-6-(cyclopropylamino)-1,3,5-triazine-2-yl] amino]-2-methylpropanenitrile} on the photochemical reactions of isolated pea (Pisum sativum L. ‘Progress #9 Dwarf’) chloroplasts were studied. Atrazine, cyanazine, and procyazine inhibited electron transport but did not uncouple photophosphorylation. The primary site of inhibition for all three herbicides was on the reducing side of photosystem II; the electron transfer step between the primary electron acceptor (Q) and the plastoquinone pool of the electron transport chain is suggested as the site of action of all three herbicides. The amount of inhibition of electron transport observed after addition of herbicide to isolated chloroplasts was time-dependent for cyanazine and procyazine but not for atrazine. This was apparently due to a slower partitioning of cyanazine and procyazine from the aqueous phase of the reaction solution into the highly hydrophobic environment within the chloroplast membrane. Treatment of the thylakoid membranes with detergent reduced the time-dependent inhibitory nature of cyanazine and procyazine, and the ability of atrazine to block electron transport. A photosystem II-dependent electron transport assay and a chlorophyll fluorescence induction assay were used to determine the inhibitory potentials of atrazine, cyanazine, and procyazine. After allowing for differences in the rate of membrane penetration, I50 values of approximately 2 × 10−7 M were determined for each of the three herbicides.

A comparison of the structure of the α-methylbenzyl analogue (MBAT) of atrazine with the Photosystem II herbicides atrazine [6-chloro-N-ethyl-N′-(1-methylethyl)-1,3,5-triazine-2,4-diamine] and diuron [N′-(3,4-dichlorophenyl)-N,N-dimethylurea] suggested that N-alkylbenzyltriazines may be a structural bridge between the triazines and the phenylureas. In the phenylureas, the addition of chlorines at the meta and/or para positions produces a marked increase in activity. Chloro-substituted derivatives of MBAT were synthesized to determine whether this structure-activity relationship also applies to the alkylaryltriazines. Addition of a chlorine to MBAT at the 4-position (CMBAT) caused a substantial increase in intrinsic activity, and a second chlorine at the 3-position (DCMBAT) caused a further increase. In direct comparisons, DCMBAT was more active in vitro than terbuthylazine [6-chloro-N-ethyl-N′-(1,1-dimethylethyl)-1,3,5-triazine-2,4-diamine], the most active chlorotriazine, and was also more active than diruon. The effects of DCMBAT were also measured on triazine-resistant pigweed (Amaranthus hybridus L. # AMACH) both in vivo and in vitro. The activity of this compound in triazine-resistant chloroplasts was intermediate between that of atrazine and diruon both in inhibition of photosynthetic electron transport and in competition for diuron binding sites, with half-maximal values falling in the micromolar range. Whole plant phytotoxicity of DCMBAT on triazine-resistant pigweed was also intermediate between that of diuron and atrazine. Since DCMBAT is a triazine with biological properties similar to that of a urea, we conclude that in a functional as well as structural sense, DCMBAT is a herbicide that is a hybrid between a triazine and a urea.

The tolerance of several field pea cultivars to glyphosate was compared in the laboratory and greenhouse, using root, foliar, and tissue culture exposure techniques. Pea cultivar response among glyphosate exposure techniques did not always agree. However, the cultivar ‘Alaska’ was consistently one of the most susceptible cultivars regardless of exposure technique. ‘Melrose’ was one of the most tolerant cultivars, especially when its roots and cells were exposed to glyphosate. The response of ‘Frogel’, ‘Glacier’, and other cultivars varied among glyphosate-exposure techniques.

Applications of glyphosate [N-(phosphonomethyl)glycine] to the leaves of pea (Pisum sativum L. ‘Burpee's Blue Bantam’), sunflower (Helianthus annuus L. ‘Mammoth’), and bean (Phaseolus vulgaris L. ‘Red Kidney’) reduced transpiration in all three species. The reduction varied with species and concentration applied. Eight-day-old bean leaves responded earlier than leaves on older plants. However, after 29 h, there was no difference in transpiration rates with seedling age, since treated leaves on both ages investigated were transpiring at about one-half the rate of control leaves. Addition of surfactant to the solution shortened the time interval between treatment and transpirational response but did not affect the final transpiration rate as compared to glyphosate without surfactant. The effect of glyphosate on transpiration was not due to changes in water availability or to the surface of the leaf that was treated. It is suggested that these results can give some insight into uptake, translocation, and mode of action of glyphosate.

Laboratory and greenhouse studies were conducted to determine the mode of action of soil- and foliar-applied UCC-C4243. Experiments demonstrated that UCC-C4243 required light for phytotoxicity, phytotoxic symptoms were similar to inhibitors of porphyrin synthesis such as acifluorfen, and UCC-C4243 potently inhibited protoporphyrinogen oxidase. Germination and emergence of field pennycress and lentil in the dark were not affected by soil-incorporated UCC-C4243 at rates more than 10 times greater than like treatments that killed all plants in the light. Soil-incorporated UCC-C4243 required light for activity and killed seedlings within 1 d after emergence; sublethal doses caused desiccation, veinal necrosis, and leaf deformation. Field pennycress and lentil were susceptible to soil-incorporated UCC-C4243 and acifluorfen in the light, but were 5 to 93 times less sensitive to the herbicides in the dark. Wheat was not affected by either herbicide in the light or dark. Injury symptoms from UCC-C4243 applied POST to redroot pigweed were similar to symptoms from diphenyl ether and bipyridinium herbicides: rapid, light-dependent chlorophyll bleaching, desiccation, and necrosis. UCC-C4243, acifluorfen-methyl, and acifluorfen acid caused light- and concentration-dependent chlorophyll bleaching and electrolyte leakage from cucumber leaf disks (I50 = 1.0, 1.8, and 4.3 μM, respectively). An inhibitor of the porphyrin synthesis pathway, 4,6-dioxoheptanoic acid, almost completely inhibited herbicide-induced electrolyte leakage. δ-Aminolevulinic acid, a tetrapyrrole precursor and stimulator of the porphyrin synthesis pathway, caused synergistic effects with each herbicide. Protoporphyrinogen oxidase from barley etioplast preparations was inhibited 50% by 40 nM UCC-C4243. Barley leaf sections treated with 100 μM UCC-C4243 accumulated protoporphyrin IX in vivo to levels > 75 times non-treated controls. These data indicate the light-requiring herbicide activity of UCC-C4243, like acifluorfen, is due to inhibition of protoporphyrinogen oxidase.

Five sulfonylurea herbicides were applied to Typic Boroll soil with pH 8.0 and 2% organic matter in the field and growth chamber to assess their residual effects on following crops. Under irrigated conditions, triasulfuron at 22 g/ha reduced the growth of alfalfa, canola, corn, lentil, pea, potato, and sugarbeet the year after application but caused no injury after two years. Alfalfa GR50 (rate required to reduce growth by 50%) values immediately after application of chlorsulfuron, triasulfuron, and metsulfuron were < 2.2 g/ha and similar in the growth chamber and field. Alfalfa and canola GR50 values immediately after thifensulfuron application in the growth chamber were ≥ 90 g/ha. Alfalfa, canola, flax, and sugarbeet GR50 for tribenuron in the growth chamber were < 4 g/ha. Crop response curves developed by growth chamber experiments could replace many of the multiple site and crop field experiments now required to develop recropping guidelines for residual herbicides.

Thiamine is a vitamin that has been shown to act as a trigger to activate plant defence and reduce pathogen and nematode infection as well as aphid settling and reproduction. We have here investigated whether thiamine treatments of seeds (i.e. seed dressing) would increase plant resistance against aphids and whether this would have different effects on a generalist than on specialist aphids. Seeds of wheat, barley, oat and pea were treated with thiamine alone or in combination with the biocontrol bacteria Pseudomonas chlororaphis MA 342 (MA 342). Plants were grown in climate chambers. The effects of seed treatment on fecundity, host acceptance and life span were studied on specialist aphids bird cherry-oat aphid (Rhopalosiphum padi L.) and pea aphid (Acyrthosiphon pisum Harris) and on the generalist green peach aphid (Myzus persicae, Sulzer). Thiamine seed treatments reduced reproduction and host acceptance of all three aphid species. The number of days to reproduction, the length of the reproductive life, the fecundity and the intrinsic rate of increase were found reduced for bird cherry-oat aphid after thiamine treatment of the cereal seeds. MA 342 did not have any effect in any of the plant-aphid combinations, except a weak decrease of pea aphid reproduction on pea. The results show that there are no differential effects of either thiamine or MA 342 seed treatments on specialist and generalist aphids and suggest that seed treatments with thiamine has a potential in aphid pest management.

A primary challenge of managing vegetable production on a small land base is the maintenance and building of soil quality. Previous studies have demonstrated the benefits of cover crops for improved soil quality; however, small growers struggle to fit cover crops into rotations. Small-scale growers with limited available land are under significant pressure to maximize their saleable yield and often work to maximize output by using intensive cropping practices that may include both early and late season crops, thus limiting the typical shoulder season windows in which cover crops can be grown. In-season living mulches may be an effective strategy to provide small-scale growers the benefits of cover crops with less land commitment than cover crops used in typical rotations. However, research on living mulches is generally not suited to small-scale organic production systems due to the typical reliance on chemical herbicide to suppress mulches. An experiment was designed with the goal of evaluating living mulch systems for space-limited organic vegetable production. In a 2-year study, four living mulch crops (buckwheat (Fagopyrum esculentum), field pea (Pisum sativum), crimson clover (Trifolium incarnatum) and medium red clover (Trifolium pratense)) and a cultivated control with no mulch cover were planted in early spring each year. Snap beans (Phaseolus vulgaris var. Tavera), transplanted bell peppers (Capsicum annuum var. Revolution), and transplanted fall broccoli (Brassica oleracea var. Imperial) were then planted directly into living mulches. During each summer growing season, living mulches and weeds were mown between-rows and hand-weeded in-row approximately every 10–14 days as needed for management. Labor times for mowing and cultivation were found to be higher in all treatments relative to the cultivated control. An inverse relationship between living mulch biomass and weed biomass was observed, demonstrating that living mulches may contribute to weed suppression. However, lower vegetable yields were seen in the living mulch treatments, most likely due to resource competition among vegetables, living mulches and weeds. High pre-existing weed seedbank and drought conditions likely increased competition and contributed to reduced vegetable yield.

Plant protease inhibitors (PI) constitute a major class of defence proteins being selected as an important strategy towards insect herbivory. With the objective of assessing the effect of PI towards larval growth and development, an artificial diet bioassay using partially purified PI obtained from peas was performed on the melon fruit fly Bactrocera cucurbitae (Coquillett). Larval growth and developmental parameters were assessed at different concentrations (namely 12.5, 25, 50, 100, 200 and 400 μg/ml) on the second-instar larvae of B. cucurbitae. Growth and survival responses determined the anti-insect potential of this PI even in its partially purified state. Larval and total development periods were found to be significantly prolonged for the larvae fed on an artificial diet incorporated with pea PI when compared with those fed with a control diet. Furthermore, when compared with the effect of the control diet (no inhibitor), the partially purified pea PI in the diet reduced larval weight gain, mean larval growth rate and food assimilated with respect to the control of the second-instar larvae tested at the same range of concentrations. The relative effectiveness of pea PI on these parameters is in agreement with the results obtained for percentage of pupation and percentage of adult emergence, as these parameters were significantly affected by the increase in the PI concentration in the artificial diet. Feeding the second-instar larvae a diet containing a range of concentrations (50, 100, 200 and 400 μg/ml) of partially purified pea PI significantly reduced the activities of digestive enzymes (trypsin, chymotrypsin, elastase and leucine aminopeptidase) and significantly affected the activities of other non-digestive enzymes (esterase, acid and alkaline phosphatases, glutathione S-transferase, superoxide dismutase and catalase).

The pea leaf weevil, Sitona lineatus L., is periodically a significant pest of pea, Pisum sativum L., in the Palouse region of northern Idaho and eastern Washington, USA. Previous on-station research demonstrated significantly greater adult pea leaf weevil colonization, immature survival, adult emergence and plant damage in conventional-tillage compared to no-tillage plots of pea. In experiments conducted during the 2006 and 2007 growing seasons, aerial and ground adult pea leaf weevil colonization of large-scale commercial pea fields under different tillage regimes in northern Idaho and eastern Washington was examined for the first time. Initial pea leaf weevil feeding damage, immature weevil densities and subsequent adult emergence from the fields were also assessed. During both years, significantly more adult pea leaf weevils were captured in conventional-tillage than in no-tillage fields during the crop establishment period in May. No-tillage soils remained wet longer in the spring and could not be planted by growers until later than conventional-tillage fields. Pea planted under conventional-tillage emerged earlier and had significantly greater feeding damage by the pea leaf weevil than no-tillage pea. Significantly, greater immature pea leaf weevil densities and subsequent adult emergence were observed in conventional-tillage than in no-tillage pea fields. Delayed development of root nodules in the cooler, moister conditions of no-tillage pea fields likely resulted in escape from attack and injury during the critical growth stages that ultimately influence yield. Results indicate that large-scale commercial no-tillage pea fields are less suitable for colonization and survival of the pea leaf weevil and suffer less weevil damage than fields under conventional tillage.

Legumes are crops that develop in cropping systems with relatively low inputs and are suitable to a more sustainable agriculture. Successful crop establishment, which is crucial for reliable plant production, depends on seed quality, environmental factors and genotypes. We studied pea and bean germination and seedling growth at various temperatures (5–40°C) and water potentials ( − 0.2 to − 1.5 MPa) using winter and spring pea and two common bean seeds produced in different conditions. The germination base temperature was − 1.1°C for pea seeds, and seeds of the winter genotype germinated more rapidly than those of the spring genotype. The base temperature for bean seed germination was 5.1–9.6°C, depending on the seed lot. The germination base water potential was about − 2 MPa for both species. The base temperatures for shoot elongation were higher (3–6°C) than those for germination. A review of the literature on other legumes confirmed that the differences in the responses of the legume seeds and seedlings to different temperatures were associated with their geographic origin. These results help understanding of pea and bean crop establishment, provide crop model parameter values and contribute to the search for genetic variability.

Dehydrin and QP47, proteins present in mature pea seeds (Pisum sativum), have been proposed to play protective roles during desiccation. To identify possible relationships between these proteins and desiccation tolerance, their tissue locations and patterns of synthesis and degradation have been examined during germination. Tissue locations were determined by immunocytochemistry using polyclonal antibodies raised against a conserved dehydrin amino acid sequence and against purified QP47. In embryonic axis and cotyledon cells, QP47 and dehydrin were distributed uniformly with no apparent nuclear or organellar specificity. Both proteins were present in 24 h-imbibed axes that had not initiated radicle growth but were completely absent from 24 h-imbibed axes that had begun to grow. The amounts of QP47 and dehydrin in embryonic axes decreased with time after the start of imbibition and were undetectable by 48 h. When germination was prevented by polyethylene glycol (PEG) or abscisic acid (ABA), both proteins remained at their original amounts. Thus, both QP47 and dehydrin disappeared coincidently with the beginning of growth and not simply as a function of the time after imbibition. QP47 persisted in cotyledons until at least 31 days into seedling growth, whereas dehydrin was not detectable in cotyledons after 7 days. Dehydrin, but not QP47, could be re-induced in pea shoots and cotyledons by dehydration. The timing of degradation of both proteins was correlated with the loss of desiccation tolerance during germination of pea axes.

Total protein content and amounts of albumin, legumin and vicilin have been determined for pea seeds from lines near-isogenic except for genes at the rugosus loci, r and rb (RR/RbRb; rr/RbRb; RR/rbrb; rr/rbrb). Seeds with the wildtype, round-seeded phenotype (RR/RbRb) had less protein on a total seed dry-weight basis than any of the wrinkled-seeded lines and this protein had a lower proportion of albumin. The lines which had recessive alleles at both r and rb loci had the highest proportion of protein and the highest proportion of albumin. The roundseeded peas possessed nearly two-fold more legumin than the double recessive line, with proportions for the two single recessive lines falling in between these extremes. Vicilin levels were similarfor all four near-isogenic lines. SDS-PAGE analysis of the isolated albumin, legumin and vicilin fractions revealed no significant differences between the four lines. Differential scanning calorimetry of protein extracts showed that all the wrinkled-seeded near-isolines possessed legumin fractions with diminished thermal stability relative to that from the roundseeded, wild-type line.

Chemically-induced mutants were also analysed for protein content and composition. These mutants have previously been shown to display great variation in starch and lipid levels. Total protein varied from 20.3% to 37.9%; however, relative proportions of albumin, legumin and vicilin were similar in all mutant lines. SDS-PAGE analysis identified two mutant pea lines which possessed a legumin A-chain of 65 000 Mr as well as the typical 45 000 Mr form. Differential scanning calorimetry of protein extracts indicated that the legumin in all mutants had lower enthalpies of denaturation than the legumin in the round-seeded parent.

The mutant pea lines possess exceptional variation with respect to starch, lipid and protein which raises opportunities for their use in the food and animal feedstuff industries.

On the premise that seed ageing may be largely a result of free-radical lipid autoxidation, a study was made of the relationship between lipid stability and longevity in seeds of soybean (Glycine max), lentil (Lens culinaris), mungbean (Vigna radiata), chickpea (Cicer arietinum), broadbean (Vicia faba), pea (Pisum sativum), lettuce (Lactuca sativa), and tomato (Lycopersicon esculentum). Seed lipids were examined for α-, γ- and δ-tocopherols, and levels of unsaturated and saturated fatty acids. Using this information, analysis was attempted using linear, multiplicative or exponential models to correlate aspects of lipid stability with seed longevity values from the published literature. No statistically significant correlations could be found between longevity and total lipid unsaturation, tocopherol levels or two protection formulae obtained from the oil chemistry literature. When values for tomato were excluded, a good correlation (r = 0.89, P = 0.007) was obtained using a multiplicative regression model for levels of linolenic acid per unit of total tocopherols in relation to longevity. Possible factors contributing to a lessening of the relationship between lipid stability and seed longevity are discussed.

The value of the garden pea (Pisum sativum) both as an experimental tool and food crop has led to an accumulated wealth of genetic variation. This paper describes how some of the variation may be harnessed for studying seed development. One can identify genotypic differences for maternal, embryonic and cellular components and the interaction between them. The impact of these components on processes such as cell division, partitioning and the deposition of storage products can then be assessed by utilizing new techniques such as immunocytochemistry and in situ hybridization.

The rugosus loci, which induce wrinkling of the seed, are examples of how major genes can be exploited to dissect seed development. Alleles at one of these loci, r, have been defined at both the molecular and biochemical levels. A much clearer picture can now be drawn of the pleiotropic nature of the genes, from their effects on starch composition to those on cellular development, which has implications for seed development in many species. The rugosus loci are the only ones known to affect seed development in peas. By analogy with cereal seeds, however, one would anticipate other mutants in the starch pathway which also affect the shape of the seed. Here, we report the induction of new rugosus-like mutants, compare them to starch mutants in other species and examine the ways in which they may help our understanding of starch biosynthesis and the regulation of seed development.

A protein named QP47 has been purified from quiescent nuclei of root meristems of Pisum sativum, and used to prepare a polyclonal antibody. Immunolocalization of this protein with fluorescent probes revealed a nuclear distribution of thread-like structures. However, the relationship between the distribution of QP47 immunofluorescence and the structural organization of the chromatin required further investigation. The decrease in content of this protein in the nuclei of embryo cells seems to be correlated with the transition from quiescence to proliferation. QP47 degradation seems to depend upon an increase in the state of its phosphorylation. This protein is not present in normally proliferating cells, or in cells whose cell proliferation has been arrested by starvation or differentiation. It is hypothesized that QP47 may be required specifically during the quiescent period for specific structural organization of the chromatin.

Interspecific competition between an introduced parasitoid species aimed at controlling a herbivorous pest species and a native parasitoid parasitising the same host may influence the success of classical biological control programmes. In Kenya, interspecific competition between an introduced and a local parasitoid on two diamondback moth populations (DBM, Plutella xylostella) was investigated on two different host plants. We tested simultaneous and delayed competition of the local parasitoid Diadegma mollipla Holmgren and its exotic congenus D. semiclausum Hellen on a newly aquired DBM host plant (snowpea) in the laboratory. Under simultaneous competition, D. mollipla produced more progeny than D. semiclausum on snowpea. A head start of D. Mollipla, of four and eight hours before its congenus was introduced, resulted in a similar number of progeny of both species. In delayed competition (time intervals of 24 h, 48 h and 72 h), progeny production was similar for both parasitoids when the time interval was 24 h, irrespective of which species parasitized first. More progeny was produced by the species which attacked first, when the time interval was greater than 24 h, although it was only significant at 72 h. Competitive abilites of both parasitoids on the new host plant differed largely between laboratory and semi-field conditions. The influence of two host plants (snowpea and cabbage) on competition was studied in the greenhouse with different host and parasitoid densities. Parasitism levels of D. semiclausum were significantly higher than those of D. mollipla, regardless of host plant, host and parasitoid densities, but progeny production of D. mollipla on snowpea was still slightly higher than on cabbage. As compared to the confinement of parasitoids and larvae to small containers, D. mollipla parasitized very few larvae in the cages. Competitive ability of the two parasitoid species tested was influenced both by the density of the searching females and by parameters related to either the host plant and/or the herbivorous hosts.

Host shifts in herbivorous insects are thought to sometimes provide enemy-free space on the novel host plant. A population of the diamondback moth Plutella xylostella (Linnaeus), an oligophagous pest on crucifers, recently shifted to sugar snap- and snowpeas (Pisum sativum) in Kenya, resulting in heavy damage to these crops. The impact of this host shift on the interaction with Diadegma mollipla (Holmgren), one of the most frequent parasitoid species attacking P. xylostella in this area, was investigated. Parasitism rates and development of two strains of D. mollipla, one reared from a cabbage-feeding strain of P. xylostella and the second from the new pea-feeding strain, changed based on the host-plant that P. xylostella fed upon, with both parasitoid strains more effective on the novel host plant. Parasitism by the cabbage-D. mollipla strain on P. xylostella infesting peas was four times higher than on P. xylostella infesting cabbage when a single plant species was present. However, when both crops were offered together, the level of parasitism dropped to the level seen when cabbage was offered alone. Diadegma mollipla developed on both hosts, but cabbage-D. mollipla had a longer total development time. Pupae of cabbage-feeding P. xylostella were significantly heavier than pupae of pea-feeding P. xylostella and parasitism had no influence on these differences. Diadegma mollipla preferred to parasitize the pea-feeding P. xylostella. Thus, the host shift by P. xylostella to a novel host plant did not necessarily provide enemy-free space, with the parasitoid species tested. The implications of these findings for the host–parasitoid relationship are discussed.