For experiments with green and brown leaves, isolations were initiated immediately after collection, with most samples processed within 29 days. All isolations were completed by 46 days after collection. Storage period did not alter results when included as a covariate in models for these experiments and was excluded from the final analyses. Isolations from leaves of the yellow leaf experiment were completed within nine days after collection, and all isolations from a single collection week were completed in one day. The presence of P. ramorum and P. gonapodyides was determined by microscopic examination of isolate morphologies directly from the isolation plates after four to five days and checked again periodically for three weeks. To test for active sporulation from colonized leaves in the microcosms, periodically a California bay leaf disc was floated as bait—either naked or in a roughly 35 mm2 mesh envelope—on the surface of the water in each microcosm for three to seven days, after which it was surface sterilized and isolations attempted from it on selective PARP-H medium. We conducted these tests of sporulation four times during the first and yellow leaf experiments, and three times during the second experiment. Additionally, we tested for sporulation periodically for up to eight weeks after all leaves had been removed from microcosms to determine if Phytophthora spores could persist in the absence of a substrate. The first experiment was initiated on 29 December 2014, but we delayed the first collection at two weeks by two additional weeks because zoospores were not detected in the microcosms until two weeks after inoculation,grow hydroponic most likely due to excessive aeration of the water during the first week.

Collections are reported according to the originally planned intervals of 2, 4, 8 and 16 weeks, with time zero being two weeks after inoculation. The final collection for the first experiment was on 1 May 2015 . Subsequent experiments proceeded as expected and the collection week reflects the period elapsed since introducing inoculum. The second experiment was initiated 28 August 2015 and concluded with the last sampling on 18 December 2015 . The yellow leaf experiment was initiated on 13 September 2015 and the final collection made on 7 January 2016 . To evaluate the colonization of leaves by each Phytophthora species in each treatment, we recorded the total number of pieces yielding P. ramorum or P. gonapodyides out of the total number of pieces sampled for each leaf. The average proportion of leaves colonized by either species was calculated for each packet from this ratio. This average leaf fraction colonized per packet was logit transformedto normalize variances, with a +0.005 correction applied to values of zero and −0.005 to values of one before transformation. The transformed average proportions colonized were analyzed in linear mixed models with the nlme package in R statistical software, version 3.3.1. Replication block and microcosm were set as random variables with microcosm nested in a block. Because Phytophthora recovery followed a non-linear trend with respect to time, we treated the collection week as a categorical variable. As one Phytophthora species occurred almost exclusively in each treatment —P. ramorum and P. gonapodyides in the treatments where they were inoculated solely and P. gonapodyides in the combined inoculations—we simplified the analysis by comparing leaf colonization by the dominant species across treatments. That is, the response variable in the model was the average fraction of leaf discs colonized by P. ramorum in P. ramorum-only treatments, and by P. gonapodyides in P. gonapodyides-only and combined Phytophthora inoculum treatments. Therefore, the main independent variables for Phytophthora leaf colonization analyses were the inoculation treatment—with non-inoculated treatments excluded—and collection week.

Leaf type and stream water type were included as independent variables in the model for the experiments where the distinctions applied. The full set of interactions were included in the models for each experiment . We verified adherence to model assumptions by the Shapiro−Wilk and Levene’s tests. We obtained P-values using the anova function in R with the sum of squares set to type III , and least square means comparisons with the lsmeans package. significance for means comparisons was determined with the default Tukey’s HSD. For leaf decomposition, we estimated a decay constant for each treatment combination in each block based on the fraction of estimated original leaf mass remaining at each collection interval. For this, we used the exponential decay equation Mt = M0 · e −kt where t is time as the number of incubation days, Mt is the fraction of leaf mass remaining at each collection interval, and M0, fraction at time zero, is set to one. Values for k were estimated using the nls function in R statistical program. The decay constants for each treatment combination were then analyzed in a mixed model using the lme function of the nlme package with inoculum, leaf and water type as independent variables and block as a random factor. For the yellow leaf experiment, only inoculum was used as an independent variable, and since replications were not blocked, an analysis of variance was performed using the aov function in R. For all experiments, we included treatments not inoculated with either Phytophthora species in the analysis to evaluate the effect of Phytophthora colonization on leaf decay. Two non-inoculated microcosms in the first experiment were contaminated with both Phytophthora species, and one non-inoculated microcosm in the second experiment became contaminated with P. ramorum, likely from a rare, undetected leaf infection. We excluded the results from these microcosms from the analysis. When P. ramorum was inoculated alone, it rapidly colonized most of the green leaf area and persisted at this level throughout the 16 weeks of incubation . It did not effectively colonize brown, senesced leaves, though it could occasionally be recovered from a few pieces of some leaves. In contrast, P. gonapodyides colonized most of the area of both green and brown leaves in microcosms where it was inoculated . However, P. gonapodyides colonized brown leaves to a significantly lesser degree than green leaves when the leaves were exposed to inoculum in separate microcosms, while there was no difference between the colonization of green and brown leaves when they were maintained in the same microcosm . In combined inoculations of both Phytophthora species, P. ramorum was unexpectedly suppressed on both leaf types and the recovery of P. gonapodyides fromthis treatment was identical to that of P. gonapodyides-only treatments . reflecting these results, in both experiments with green and brown leaves, the interaction of inoculation and leaf type was highly significant .

In the experiment with yellow leaves, P. ramorum, when inoculated alone, colonized most of the leaf area and persisted at this level throughout the experiment, similar to the result with green leaves in other experiments . In combined P. ramorum and P. gonapodyides inoculations, P. ramorum was once again completely suppressed and P. gonapodyides colonized yellow leaves almost completely, at levels similar to its colonization of green leaves in both other experiments . The colonization of yellow leaves by P. gonapodyides in combined Phytophthora inoculum treatments was significantly higher than that by P. ramorum in P. ramorum-only inoculated treatments in this experiment, though both species colonized more than 70% of the leaf area. Thus, in the experiment with yellow leaves, only the effect of Phytophthora inoculation was significant . In all experiments,mobile grow rack both Phytophthora species colonized leaves rapidly, in most cases reaching maximum levels by four weeks, and persisted at these levels throughout the 16 weeks experimental duration. A slight increase in the level of colonization by both Phytophthora species was apparent in many cases from two to four weeks, though for brown leaves maintained in separate microcosms in the second experiment, levels appeared to actually decline after the second week. This contrast is reflected in the significant interaction of leaf type and collection week for this experiment .The isolation of P. ramorum or P. gonapodyides from California bay leaf disc baits deployed on the water surface in microcosms indicated the presence of zoospores. The sum of successful bait isolations for each treatment across the five replication blocks in the green and brown leaf experiments, and across four replications in the yellow leaf experiment, are presented in Tables 1–3. The recovery of P. gonapodyides from P. gonapodyides-only and combined Phytophthora inoculation treatments was from nearly 100% of baits throughout the duration of all experiments. The recovery of P. ramorum was more erratic, ranging from 40% to 90% of baits during the experiments with green and brown leaves. However, P. ramorum recovery from baits in microcosms that included green leaves and sterile rather than non-sterilized water was closer to 100%, excepting the second baiting of the second experiment, when P. ramorum was not recovered from most microcosms. Phytophthora ramorum was also rarely recovered by baiting from microcosms in the second experiment with only brown leaves, especially when excluding the first baiting, which was done a few days after inoculation. Consistent with this, brown leaves were colonized at very low levels by P. ramorum. Nevertheless, at 14 weeks, P. ramorum could still be recovered from several of these microcosms . Phytophthora ramorum was recovered somewhat more frequently from sterile than non-sterile stream water treatment. Such an effect was not apparent for P. gonapodyides. Both Phytophthora species were recovered at nearly 100% from baits throughout the yellow leaf experiment which used sterile dilute nutrient solution only . Additionally, we baited microcosms for weeks after all leaf packets had been collected to see how long spores may persist in the absence of leaves.

Phytophthora ramorum could be recovered from a few microcosms up to six weeks after all leaves were removed, but its frequency generally diminished rapidly. In contrast, P. gonapodyides could be recovered for up to 12 weeks after all leaves had been removed from microcosms, and was relatively frequent even six weeks after leaves were removed in the second experiment.In the experiment with green and brown leaves maintained in the same microcosm, only green leaves in microcosms with no Phytophthora inoculum decomposed at a slower rate than all other treatments . In fact, on average, they did not lose significant biomass throughout the 16 weeks. Leaves in all other treatments, including brown leaves in non-inoculated microcosms, decomposed at similar rates . The interaction of leaf type and Phytophthora inoculation was, therefore, a highly significant predictor in the model . Estimated decay constants are listed in Table S5. In the experiment where green and brown leaves were maintained in different microcosms, green leaves in microcosms with Phytophthora inoculum decomposed faster than all other treatments . In this experiment, all treatments with brown leaves and green leaf treatments with no Phytophthora decomposed at similar rates. Notably, in contrast to the other experiment, green leaves in non-inoculated treatments in this experiment did decompose over the 16 weeks, ultimately achieving a similar level of biomass loss as green leaf treatments with Phytophthora inoculum. Nevertheless, reflecting the difference in decomposition rate for green leaves in inoculated and noninoculated treatments, the effect of the leaf type by Phytophthora inoculum interaction was significant in the model . Estimated decay constants are listed In the experiment with yellow leaves only, leaves in the non-inoculated treatment decomposed at a slightly but significantly lower rate than Phytophthora-inoculated treatments , of which the decomposition rates were not statistically different from one another . The decomposition rate of yellow leaves in Phytophthora-inoculated treatments was similar to that of green leaves in Phytophthora-inoculated treatments of the second experiment with which it was essentially concurrent, though the results of the different experiments were not statistically compared.The goal of these experiments, broadly, was to better understand how the previously observed differences in trophic specialization between P. ramorum and P. gonapodyides, the latter as a representative of stream-resident clade 6 Phytophthora species, affected their ability to utilize different kinds of leaf litter available in streams. More specifically, we sought to determine if the previously observed decline of P. ramorum in green leaves decomposing in streams was due to competitive displacement by saprotrophic organisms or due to an intrinsic inability of this pathogenic species to persist on colonized but decomposing leaf tissue, and to discover if the observed specialization of each species as pathogen or saprotroph would be consequential for the colonization of senescent or fully senesced leaves, a factor that has important implications regarding the prevalence of suitable leaf litter substrate for these organisms.