Influence Of Cell Density And Phase Variants Of Bacterial Symbiont Biology Essay
Entomopathogenic nematodes of the genus Steinernema (Nematoda: Rhabditida) are symbiotically associated with enterobacteria of the genus Xenorhabdus, member of the gamma subclass of the Proteobacteria (Ehlers et al. 1988). The bacto-helminthic complexes are safe biocontrol agents (Ehlers 2003) used to manage several insect pests in ornamentals and food crops (e.g., Grewal et al. 2005, Toepfer et al. 2008). Like other rhabditid nematodes, Steinernema spp. have a developmentally arrested third juvenile stage, called dauer juvenile (DJ), which is the only free-living stage. The term dauer (German for enduring) describes a morphologically distinct stage, formed as a response to depleting food sources and adverse environmental conditions. DJs are well adapted for long term survival in the soil environment (Susurluk and Ehlers 2008). Each DJ carries some hundred cells of its specific symbiont in a pocket in the anterior part of its intestine (Bird and Akhurst 1983). The DJ is actively searching for suitable insect hosts (Rasmann et al. 2005) and penetrates into the insect´s haemocoel. Inside the insect, the complex encounters suitable conditions for reproduction and the DJs leave the DJ stage (Han and Ehlers 2000). In accordance to the rhabditid nematode Caenorhabditis elegans, the exit from the DJ stage is called ”recovery” (Golden and Riddle 1982, Strauch et al. 1994). Recovery is a response to food signals in the haemolymph. During recovery the bacteria are released, which accelerate the killing of the host (Han and Ehlers 2000). The nematodes feed on the symbiont cells, grow to adults and produce DJ offspring, which leave the insect’s cadaver on the search for new hosts (Ciche et al. 2006).
Xenorhabdus spp. produce phase variants with different phenotypes (Akhurst 1980). Secondary forms are detected during the stationary growth phase in cultures, which have been inoculated with primary phase. Primary cells absorb bromothymol blue from NBTA agar plates. Secondary colonies cannot absorb the dyes from agar media (Akhurst 1980). Primary cells carry inclusion proteins, produce antibiotic substances, lipase, phospholipase and protease. Secondary variants loose these characters (Boemare 2002). Nematodes grown on secondary phase yield significantly less DJs than if cultured on primary phase bacteria, particularly in liquid cultures (Han and Ehlers 2001).
For biocontrol purposes nematodes are produced in industry scale bioreactors. Much information on liquid culture production of the related biocontrol nematode Heterorhabditis bacteriophora Poinar 1979 has been published (Ehlers 2001). Artificial liquid media are pre-incubated with the symbiotic bacteria for approximately one day before DJs are inoculated. In vitro production is possible due to pre-culture of the symbiotic bacterium Photorhabdus luminescens, which excretes food signals into the medium and trigger DJ recovery (Aumann and Ehlers 2001). Artificial growth media lack any kind of food signal. When Strauch and Ehlers (1998) transferred DJs into culture filtrates of P. luminescens, the highest recovery was recorded in samples taken during the late logarithmic growth phase. Ehlers et al. (1998) reported that low DJ recovery leads to a non-synchronous population development of H. bacteriophora, prolonged culture duration and unstable yields. In order to reach the necessary density also at low DJ recovery, higher numbers of DJs must be inoculated, which makes necessary additional scale-up steps, thus increasing the production costs (Ehlers et al. 1998). Consequently, the key to a successful and cost-effective liquid culture process is through the management of DJ recovery. These conditions have not yet been investigated in species of the genus Steinernema.
Under natural conditions, Heterorhabditis DJs recover after invasion of the insect´s haemocoel as a response to yet unknown food signals in the haemolymph. In haemolymph drops obtained by bleeding the insects, DJ recovery was much reduced (Strauch and Ehlers 1998). Whether steinernematids recover in insect blood serum was investigated in this study as well as the influence of host penetration on DJ recovery and evaluate the influence of Xenorhabdus cell density and their phase variants on nematode DJ recovery. This study also was performed to compare results obtained with Heterorhabditis DJ recovery with those from the the genus Steinernema, which is believed to have evolved the symbiotic relation to entomopathogenic bacteria independently (Sudhaus 1993).
Material and methods
Xenorhabditis spp. were isolated from nematode-infested insect larvae of Galleria mellonella according to Ehlers et al. (1990). Xenorhabdus nematophila was isolated from the All strain of S. carpocapsae, X. bovienii from the Sylt strain of S. feltiae. Symbiotic bacteria were then propagated in YS broth in 100 ml Erlenmeyer flasks fill with 30 ml medium containing (in g l-1) 5 NaCl, 5 yeast extract, 0.5 NH4PO4, 0.5 K2PO4, 0.2 MgSO4 x 7 H2O (Merck) at 180 rpm and 25°C. Bacterial cultures of 1010 cells ml-1 were transferred into 2 ml Eppendorf tubes and mixed with glycerol to obtain a 15% solution and stored at -20oC until use.
Steinernema carpocapsae (All strain) and S. feltiae (EN02 hybrid strain) were used. Monoxenic cultures were established according to Lunau et al. (1993). The DJ inoculum was produced according to Ehlers et al. (1998) in liquid medium (LM) containing (in g l-1) 15 yeast extract (Merck), 20 soy flour, 6 lecithin (Cargill), 40 vegetable oil (Raiffeisen), 4 NaCl, 0.35 KCl, 0.15 CaCl2, 0.1 MgSO4 (Merck). Media were adjusted to pH 7.0 for S. carpocapsae and 6.5 for S. feltiae with 4M KOH (Merck). LM was inoculated with the symbiotic bacteria of the specific nematode. Therefore bacterial stocks were thawed and propagated at 25°C in YS broth until a cell density of 109 cells ml-1 was reached. LM was inoculated at 1% with the YS bacterial culture and incubated for 36 h at 25°C. Then DJs from monoxenic cultures were inoculated at 3,000 DJs ml-1 and incubated at 25°C until harvest at 15 days after inoculation of the symbiotic bacteria.
Nematodes were collected by centrifugation (2,500 rpm, 10 min). To remove culture debris and bacteria, DJs were washed in Ringer’s solution (Merck) and centrifuged again. Then active DJs were separated from inactive ones by suspending the DJs over a 30 µm sieve in water and collecting DJs, which had passed the sieve after 4-h in sterile Ringer’s solution. DJs were stored at 4oC until use. All steps were performed under sterile conditions.
Establishment of secondary phase bacteria
The shift to secondary phase bacteria was induced by exposing the cultures to stress conditions according to Krasomil-Osterfeld (1995). X. nematophila was cultured for 20 days at 100 rpm and 25oC under osmotic stress in YS broth without NaCl. In order to identify secondary colonies, samples were streaked onto NBTA containing (in g l-1) 10 trypcase soy broth (TSB) (bioMerieux), 15 bacto-agar (Difco), 0.5 bromothymol blue (Sigma), supplemented after cooling to 35°C with 4 ml l-1 of filter-sterilized 2,3,5-triphenyl-tetrazoliumchloride (Merck). Secondary phase bacteria produce red colonies on NBTA, whereas primary absorb bromothymol blue (Akhurst 1980). A single secondary colony was selected and sub-cultured again under osmotic stress. After four selection rounds, pure secondary phase bacteria were obtained. They were then once propagated in normal YS at 180 rpm and afterwards checked for phase stability. Secondary phase of X. bovienii was obtained from 2 month old monoxenic flask cultures of S. feltiae, which had been started with primary cells. The secondary phase was stable even without culture under osmotic stress. Storage was also done in 15% glycerol stocks at -20oC.
Phenotypic characterization of secondary phase of X. nematophila and X. bovienii
Further secondary phase characters were checked to support identification by dye absorbtion. Whereas primary bacteria are mobile, swarming on soft agar, and producing antibiotic activity against Bacillus cereus in overlay cultures, the secondary bacteria loose these characters (Boemare and Akhurst 1988). Swarming was checked after culturing primary and secondary in parallel culture plates for 72 h at 25°C on soft agar containing (in g l-1) 15 TSB and 8 agar (Difco). For testing the antibiotic activity, a YS broth culture of B. cereus was mixed at 5% into sterile TSB supplemented with 15 g l-1 agar after cooling to 50oC. Primary and secondary cultures of X. nematophila and X. bovienii were incubated in YS broth until a cell density of 1010 cells ml-1 was reached. Then 1.5 ml samples were centrifuged at 14,000 rpm for 10 min and the supernatant filter-sterilized through a 0.22 µm pore size membrane (Milipore). Ten µl of each filtrate was then inoculated onto the agar containing B. cereus. Growth inhibition indicates the presence of primary cells.
Effect of insect penetration on DJ recovery
In liquid culture production DJ do not undergo host penetration prior to recovery. In order to determine whether host insect penetration affects recovery, DJs were either injected into an insect host or left to penetrate into the insect. Single last instars of the lepidopteran insect Galleria mellonella were exposed to 50 DJs of S. carpocapsae in 10 µl Ringer’s solution on a wet filter paper disc inside a cell well of a 24-cell well plate (Greiner). For S. feltiae invasion, G. mellonella larvae were buried into 10% moist sand in the well and also exposed to 50 DJs. At 6 h post inoculation (hpi), the infected larvae were collected from cell wells and transferred into plastic boxes covered with wet filter paper and maintained at 25oC until recovery was assessed.
For DJ injection into the hemocoel of G. mellonella larvae, inoculum DJ density was adjusted at a density of 5,000 DJs ml-1 in sterile Ringer’s solution. Final instar G. mellonella larvae were injected with 10 µl of the DJ suspension to achieve approximately 50 DJs per insect. Injection was performed with a microsyringe (ITO Corp, Japan). Recovery was observed at 6, 24 and 48 hpi. Twelve inoculated larvae were dissected in Petri dishes filled with Ringer’s solution and the nematodes were collected into counting chambers (Cellstar, Germany) to check the recovery under an inverted microscope (Zeiss, Germany). Dauer juveniles were distinguished from recovered juveniles by morphological characters according to Strauch and Ehlers (1998). The head region of DJs is tapered, whereas recovered juveniles have a broad head region and the mouth is opened. The experiment was repeated twice.
Effect of insect serum from G. mellonella larvae on DJ recovery
G. mellonella larvae were collected and surface sterilized by washing them in 70% ethanol. The hemolymph was collected from a proleg into a 2.0 ml cap, which contained 0.5% phenylthiourea-acetone (Sigma) at 1/10 of the total volume to avoid melanization of the haemolymph (Arakawa 1995). Collected hemolymph was stored at -20oC until use. Before the experiment, the stock was thawed and purified by centrifugation and filtration with a 0.2 µm filter. Then ampicillin and streptomycin sulphate (Merck) were added each at 0.1% (w/v) to avoid any activity of symbiotic bacteria and the sterile serum was incubated for 24-h. All handling was carried out under sterile condition. 500 µl haemolymph was added to each well and 500 DJs per well were inoculated. The cultures were then incubated on the shaker at 180 rpm and 25oC. Recovery was assessed from at least 50 juveniles at 24 and 48 hpi. DJ recovery in haemolymph was compared with the DJ recovery in living G. mellonella insect inoculated without penetration.
Influence of bacterial cell density on DJ recovery
Bacterial symbionts were cultured in LM in shaken flasks for 36 h at 180 rpm and 25°C. Cell density was counted using a Thoma chamber. Then the bacterial suspension was centrifuged at 14,000 rpm for 30 min at 4oC and the supernatant filter sterilized (0.2 µm, Millipore). Dilutions were produced by adding sterile supernatant to the bacterial suspension to obtain cell density of 108, 109 and 1010 cells ml-1. To stop bacterial growth and avoid contamination, ampicillin and streptomycin sulphate were added as described above to the bacterial suspensions, supernatant or Ringer´s solution at 24 h prior to DJ inoculation. Cell wells of 24 well plates (Greiner) were each filled with 500µl bacterial suspension, culture supernatant or Ringer´s solution and 500 DJs from monoxenic culture per cell well. The plates were then incubated on a shaker at 180 rpm and 25oC. Every 2 h at least 50 DJs were removed and the recovery was assessed according to Strauch and Ehlers (1998). Each bacterial density, supernatant and control in Ringer´s solution was tested in three replicates (cell wells) and the experiment was repeated twice with different nematode batches.
Influence of bacterial cell density on development after recovery
The same experiment was repeated without testing supernatant and Ringer´s solution. In order to test the influence of the bacterial density on recovery and nematode development, the number of DJs, recovered DJs and adults was assessed at 12, 24, 36, 48 and 60 h after nematode inoculation. Each time a minimum of 50 individuals was observed in each variant. Female nematodes were identified based on the presence of a vulva and males on the presence of the spicula. Each bacterial density was tested in three replicates and the experiment was repeated twice with different nematode batches.
Influence of bacterial phase variant on DJ recovery
Primary and secondary phase bacteria were propagated in YS broth for 36-h. Bacterial cell free supernatant was prepared from each phase variant and suspensions adjusted to a density of 1010 cells ml-1 with the supernatant supplemented with the antibiotics. Then cell wells were filled with the suspension and nematodes as described above. In these experiments, the same amount of wells was filled with the bacterial suspension without the addition of antibiotics in order to investigate the influence of the antibiotics on DJ recovery. Recovery was observed 48 h after nematode inoculation. A minimum of 50 nematodes was observed per variant. The experiments were also done in three replicates and repeated twice.
DJ recovery after different exposure periods in bacterial cell suspension
In order to investigate how long the DJs need to be exposed to the bacterial food signal for induction of the recovery process, the same experiment was repeated, but only at a cell density of 1010 cells ml-1. Then 50 DJs were removed from the suspension at 0, 0.5, 2, 4, 6, 8, 10, 12, 24 and 36 h after inoculation, washed in sterile Ringer’s solution and then transferred to cell well containing 500 µl Ringer’s solution supplemented with the antibiotics as described above. Recovery of all 50 individuals was assessed at 48 hpi. The experiment was also done in three replicates and repeated twice. The time period, at which 50% of the inoculated DJs had recovered (RT50) was calculated by Probit analysis (Finney, 1971).
Recovery data are presented in percent. Two pair comparison was analysed by the Chi2 test. Recovery data among multiple variants were converted to arcsin values and then mean for different variants were analysed for significant differences by ANOVA and the Tukey HSD test. The recovery data over time were analysed and statistically compared using the trapezoidal rule A=(tn+1-tn)(rn+rn+1)/2 (t: time after DJ inoculation, n: sample no., rn recovery after tn) estimating the integral of the curve. The higher the area under the curve the stronger is the effect of the variant on the recovery induction. Means for different variants were analysed for significant differences by ANOVA and the Tukey HSD test. For evaluation of the DJ synchronization index the regression of the recovery data over time was calculated and the indices (slopes of graphs) were compared for significant differences by ANOVA and the Tukey HSD test.
Effect of insect penetration on DJ recovery
Recovery of DJs in insect haemocoel is presented in Table 1. At 6 and 24 hpi recovery of injected S. carpocapsae DJs was higher than the recovery of the DJ, which had undergone host searching and penetration. At 48 hpi the recovery of S. carpocapsae and S. feltiae DJs reached approximately 100% regardless which way they reached the haemocoel. The delay in recovery of DJs which naturally penetrated the insect is due to the time needed to reach the host insect and advance to the haemocoel through penetration of the integument, the wall of the tracheal system of the intestine. Thus host penetration is not a pre-requisite for DJ recovery nor is the penetration process resulting in a higher recovery of the DJs.
Influence of antibiotics on DJ recovery
An influence of the antibiotics ampicillin or streptomycin sulphate on DJ recovery can be excluded, since no significant difference of DJ recovery was recorded 48 hpi in wells with or without antibiotic treatment (data not shown).
Recovery in serum of Galleria mellonella
Of S. carpocapsae DJs 88.6% and 96.9% recovered in the sterile serum at 24 and 48 hpi, respectively. In contrast, only 22.4% and 31.0% of S. feltiae DJs recovered at 24 and 48 hpi, respectively (Tab. 2), indicating that serum of this lepidopteran insect is less well suited for recovery of S. feltiae.
Influence of bacterial cell density and bacterial culture supernatant
The results on the influence of the cell density of the nematode’s specific bacterial symbionts are presented in Figure 1. Significant differences were recorded for both, S. carpocapsae (ANOVA, F = 60.6; df = 4, 41; P < 0.0001) and S. feltiae (ANOVA, F = 40.7; df = 4, 34; P < 0.0001). Ringer’s solution obviously lacks any kind of food signal as no recovery was recorded for both nematode species even after 12 hpi. Both species recovered in bacterial culture supernatant, indicating that both, X. nematophila and X. bovienii excrete food signals into the culture medium. S. carpocapsae reached 44% recovery in supernatant, which was lower than in any of the bacterial cell suspensions. S. feltiae reached 20% DJ recovery at 12 hpi in supernatant with no significant difference to the cell densities 108 (Tukey HSD, P = 0.804) and 109 ml-1 (P = 0.5). Maximum recovery was recorded at cell densities of 1010 ml-1 for both species. The reduction of the cell density to 109 ml-1 did not result in a significant drop in DJ recovery in S. carpocapsae (P = 0.384). Only the reduction to a density of 108 ml-1 caused a significant reduction in recovery to 70% at 12 hpi (P = 0.026). In contrast, the reduction of the cell density of X. bovienii to 109 and 108 ml-1 caused a significant reduction of the recovery in S. feltiae to 39% and 18% (both P < 0.0001) at 12 hpi, respectively. The response to the bacterial food signal is more rapid in S. carpocapsae than in S. feltiae (Fig. 1). The DJ synchronization index was calculated by regression of the recovery data over time until 8 h after inoculation of the DJs into the bacterial suspension and supernatant (Table 3). The better the synchronization of the DJ recovery the higher is the index.
The faster progress of the recovery in higher cell densities was confirmed by observations on exsheathment and feeding of S. carpocapsae. The step introducing the recovery process in DJs is the exsheathment from the cuticle of the prior juvenile stage, the pre-dauer J2d stage. In X. nematophila cell suspension of 1010 cells ml-1 the first S. carpocapsae juveniles shed the outer cuticle at 8 hpi, but no exsheathment was observed in 109 and 108 cells ml-1 suspensions until 12 hpi. In 1010 cells ml-1 the pharyngeal pumping, indicating the start of the feeding, was observed at 8 hpi, while no feeding was observed in 109 and 108 cells ml-1 bacterial cell suspension until 12 hpi.
Influence of bacterial cell density on nematode development after recovery
As recorded in the previous experiments, the recovery of both nematode species was again highest at the highest bacterial density and recovery of S. carpocapsae was less affected by the decreasing cell density than S. feltiae (Fig. 2). In both species the recovery increased until 24 hpi and then was constant in S. carpocapsae and at 1010 cells ml-1 in S. feltiae, but continued to increase from 31% at 12 hpi to 77% at 60 hpi at 109 cell ml-1 and from 20% to 57% at 108 cell ml-1 in S. feltiae. The bacterial cell density also affected the development of recovered juveniles to adults. In the population of S. carpocapsae incubated at a bacterial cell density of 1010 and 109 cells ml-1 (Fig. 2 A, B) the adult population was observed already at 36 hpi, while in 108 cells ml-1 the first adults occurred at 48 hpi (Fig. 2 C). The S. feltiae adult population was first observed at 24 hpi at 1010 cells ml-1 (Fig. 2 D), 36 hpi in 109 cells ml-1 (Fig. 2 E) and 48 hpi in 108 cells ml-1 (Fig. 2 F). Where the percentage of adults in S. carpocapsae reached 86.4%, 52.4% and 11.6% at 1010, 109 and 108 cells ml-1 at 60 hpi, respectively, the percentage reached only 71.2%, 19.4% and 8.5% in S. feltiae, respectively. There was no major difference in sex ratio between different bacterial cell densities in S. carpocapsae, whereas in S. feltiae more females than males were recorded at higher cell density.
Influence of bacterial phase variation on DJ recovery
Recovery of both species, S. carpocapsae and S. feltiae, was reduced in secondary cells and also in supernatant of secondary cells (Fig. 3A, B). Other than in Ringer’s solution, in which no recovery was recorded (Fig. 1), in medium YS broth a low recovery of approximately 10% was observed. Recovery in supernatants was always lower than in cell suspensions, confirming the results obtained in the first experiments (Fig. 1).
Time scale of DJ recovery
When DJs were exposed to 1010 cells ml-1 over a period of 2 to 36 h and then transferred to Ringer’s solution, the response of S. carpocapsae DJs to the food signal was almost immediate with a significant increase over time (ANOVA, F = 43.726; df = 7, 70; P < 0.0001). Significant steps were recorded between 2 and 4 h (Tukey HSD, P = 0.001) and 4 and 6 h (Tukey HSD, P = 0.004) exposure, followed by a continuous, but small increase to 84% recovery of DJs exposed for 36 h to the bacterial suspension (Fig. 4 A). In S. feltiae (ANOVA, F = 27.248; df = 7, 70; P < 0.0001) the first significant increase in recovery is after exposure of 8 to 12 h (Tukey HSD, P = 0.0003) and again from 12 to 24 h (Tukey HSD, P = 0.003). This difference between the two species is obvious also when the RT50 is compared with 6.6 h exposure for S. carpocapsae and 17.1 h for S. feltiae.
In liquid culture mass production the recovery of inoculated DJs is the first step in the nematode´s life cycle determining success or failure of the production process (Ehlers, 2001). Low DJ recovery of heterorhabditid nematodes results in low adult density at the moment of off-spring production. As a consequence many of the offspring do not develop to DJs, instead more individuals develop to adults (Strauch et al. 1994) resulting in two-generation processes and a longer process duration (Ehlers et al. 1998). A non-synchronous population development due to a delay in recovery has the same consequences. In heterorhabditid nematodes Strauch and Ehlers (1998) reported a mean recovery in their symbiont´s liquid culture for two H. megidis strains at 23 and 45% and for H. bacteriophora at 38%, ranging between 0 and 81%. This high variability was not observed in the two steinernematid species. Recovery recorded in these experiments usually was between 60 and 90% at higher bacterial density. Thus variable recovery in steinernematids is less of a problem compared with heterorhabditids. Compared to heterorhabditid nematodes, which recover in liquid culture over a period of several days (Strauch and Ehlers 1998; Johnigk et al. 2004), DJs of the two steinernematid species recover within 24 h. In S. carpocapsae the DJ recovery is even less influenced by the bacterial cell density than observed for S. feltiae. For the latter nematode a high cell density is necessary before DJ inoculation in order to obtain high DJ recovery.
Another reason why high DJ recovery is a pre-requisite for successful liquid culture production of heterorhabditid, but less so for steinernematids is related to the different copulation behaviour of nematodes of the two genera. Adult steinernematid off-spring in the F1 generation can copulate and multiply in liquid culture, whereas males of the F1 generation in heterorhabditids cannot attach to the female. Fertilization is impossible and amphimictic adults thus cannot multiply (Strauch et al. 1994). Steinernematides can compensate for low DJ recovery and immediately utilize the remaining resources by development of a second adult generation. Second generation adults of heterorhabditids produce no off-spring. Instead reproductive second generation is developing from few F1 DJs, which recover at low nematode density (result of low recovery) and reproduce as self-fertilizing hermaphrodites (Johnigk and Ehlers 1999a, b). This developmental process takes longer and is often related to losses in the F1 DJ populations during the development of the F2 DJs.
Like the symbionts Photorhabdus spp. of the nematode Heterorhabditis spp. (Strauch and Ehlers 1998), X. nematophila and X. bovienii also excrete a food signal into the medium, however, the recovery is reduced in the superantant, whereas in heterorhabditids no difference was recorded between cell suspension and cell-free culture supernatant (Strauch and Ehlers 1998). Either the Xenorhabdus food signal is instable or living cells as such trigger recovery or the cells continue to produce food signal despite the presence of antibiotics. An effect of the antibiotics on the recovery was excluded as was the influence of the penetration process on the DJ recovery. Like in Heterorhabditis the recovery of the two steinernematids inside the insects was always 100%. But other than recorded for heterorhabditid species, S. carpocapsae also recovered well in sterile insect serum produced from the heamolymph of G. mellonella, whereas the response of S. felitae was much reduced (Tab. 2). Possibly heterorhabditids and S. feltiae DJs respond to an instable component in the insect´s haemolymph, which is lost during preparation of the serum.
Dauer juvenile recovery of both Steinernema spp. was higher in primary phase bacterial cells than in secondary cells, which might be one reason, why cultures started with seondary phase symbionts most often fail (Ehlers et al. 1990; Han and Ehlers, 2001). The observation that recovery was reduced in culture supernatants was confirmed also for the secondary supernatant.
Another important observation resulting from these experiments was that not only the recovery is positively related to the cell density of the bacterial symbiont Xenorhabdus spp. but also the time necessary for development to adults, which is retarded in lower cell denisties in both nematode species tested. The dependence on the speed of the nematode development implies the need for high density bacterial precultures prior to nematode inoculation.
The scholarship to the first author by the German Academic Exchange Service (www.daad.de) is gratefully acknowledged. Thanks are also due to Olaf Strauch for help with data evaluation and to Friederike Bernsdorff for technical support.
If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please click on the link below to request removal: