Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
1. The Janzen–Connell (J–C) Model proposes that host-specific enemies maintain high tree species diversity by reducing seedling performance near conspecific adults and promoting replacement by heterospecific seedlings. Support for this model often comes from decreased performance for a species at near versus far distances from conspecific adults. However, the relative success of conspecific versus heterospecific seedlings recruiting under a given tree species is a critical, but untested, component of the J–C Model.
2. In a shade-house experiment, we tested plant–soil feedbacks as a J–C mechanism in six tropical tree species. We assessed effects of conspecific versus heterospecific cultured soil extracts on seedling performance for each species, and we compared performance of conspecific versus heterospecific seedlings grown with soil extract cultured by a particular tree species. Additionally, we tested whether soil microbes were creating these plant–soil feedbacks and whether low light increased species vulnerability to pathogens.
3. Among 30 potential comparisons of survival and mass for seedlings grown in conspecific versus heterospecific soil extracts, survival decreased in seven and increased in two, whereas mass decreased in 13 and increased in 1. To integrate survival and growth, we also examined seedling performance [(mean total mass × mean survival time)/(days of experiment)], which was lower in 16 and higher in 2 of 30 comparisons between seedlings grown with soil extract cultured by conspecific versus heterospecific individuals. Based on performance within a soil extract, conspecific seedlings were disadvantaged in 15 and favoured in 7 of 30 cases relative to heterospecific seedlings.
4. Species pairwise interactions of soil modification and seedling performance occurred regardless of sterilization, suggesting chemical mediation. Microbes lacked host-specificity and reduced performance regardless of extract source and irradiance.
5.Synthesis. These results, along with parallel research in temperate forests, suggest that plant–soil feedbacks are an important component of seedling dynamics in both ecosystems. However, negative conspecific feedbacks were more prevalent in tropical than temperate species. Thus, negative plant–soil feedbacks appear to facilitate species coexistence via negative distance-dependent processes in tropical but not temperate forests, but the feedbacks were mediated through chemical effects rather than through natural enemies as expected under the J–C Model.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Identifying mechanisms that maintain species richness is a central question in plant community ecology because competitively dominant species are expected to exclude less competitive species (Gause 1934). Janzen (1970) and Connell (1971) hypothesized that competitive exclusion could be precluded by host-specific enemies that reduce seed and/or seedling survivorship near conspecific adults and/or at high conspecific densities. The implicit assumption of the Janzen–Connell (J–C) Model is that non-competitive distance- or density-dependent (NCDD) mortality would favour establishment of heterospecific individuals, thus promoting species coexistence.
The J–C Model argues that NCDD processes are less pronounced in temperate than tropical forests because temperate forests have a less diverse tree community, lower rainfall and greater seasonality, which could result in an overall lower abundance of enemies and disproportionately fewer specialist enemies. Among the few studies that have been conducted in temperate forests, there is accumulating evidence of NCDD effects on seedling mortality and growth (Packer & Clay 2000; Hille Ris Lambers, Clark & Beckage 2002; Packer & Clay 2003; McCarthy-Neumann & Kobe 2010). However, we are not aware of any studies that use parallel experiments in both temperate and tropical forests, which would enable more direct comparisons.
Negative feedback, whereby individual plants ‘culture’ the local soil microbial community in which they grow to the detriment of themselves and other conspecific individuals (van der Putten, Van Dijk & Peters 1993; Mills & Bever 1998; Klironomos 2002; Bezemer et al. 2006; Casper & Castelli 2007; Kardol et al. 2007), may be an important mechanism that could create NCDD mortality and/or reduced growth. Natural enemies such as soil pathogens are probably effective agents in creating this type of feedback, since many of these pathogens show host specialization, short generation times, high fecundity, long persistence in soil and more limited dispersal than their hosts (Gilbert 2002). In a prior study, we found that a majority of 21 tropical tree species experienced reductions in seedling survival and/or growth when inoculated with non-sterile versus sterile soil extract cultured by conspecific adults and seedlings (McCarthy-Neumann & Kobe 2008). However, this study did not address the relative impact of conspecific versus heterospecific interactions, and thus host-specificity of these feedbacks is unknown.
We examined seedling survival and growth responses of six tropical tree species to species culturing the soil, presence of microbial pathogens and light level. We used extracts filtered from soils that had been cultured by different adult tree species with further culturing by a high density of seedlings of the same species in the shade house to test the net effects of soil biota and chemical factors. Extracts contained bacteria and pathogenic fungal spores <20 μm and water soluble chemicals. Specifically, we tested: (H1) soil extracts cultured by conspecific adults reduce seedling survival and/or growth more than soils cultured by heterospecific adults; (H2) heterospecific seedling survival and/or growth are favoured relative to conspecific seedlings in soil extracts cultured by a particular tree species; (H3) sterilization of soil extract decreases negative effects of conspecific cultured soils because of elimination of soil pathogens; and (H4) higher irradiance reduces negative effects of soil microorganisms. Following convention, we compare within-species seedling performance in conspecific and heterospecific soils (H1), but we also introduce a novel comparison: the relative performance of conspecific versus heterospecific seedlings in a given soil source, analogous to seedlings contesting a space occupied by a mature tree (H2). The latter is important in evaluating the potential for NCDD processes to promote species coexistence. Additionally, by comparing results here with a parallel study (McCarthy-Neumann & Kobe 2010), we assess whether tropical species exhibit greater sensitivity to NCDD processes than temperate species.
Materials and methods
Overview of methods
To investigate plant–soil feedbacks as a mechanism for NCDD, we derived microbial extract from areas predicted to have the strongest negative effect, i.e. near adults and at high seedling density. Thus, we collected soil near adults of each of our study species, further cultured the soil in the shadehouse with a high density of seedlings of the same species, extracted microbial inoculum from cultured soil and inoculated experimental seedlings with these varying soil extract treatments (Appendix S1 in Supporting Information).
To test the hypotheses, we assessed seedling survivorship and growth responses for six tropical tree species (Apeiba membranacea, Colubrina spinosa, Iriartea deltoidea, Pentaclethra macroloba, Prestoea decurrens and Virola koschnyi) to extracts taken from soils that had been cultured by each of the six species. These species were selected because they vary in abundance class, seedling shade tolerance and seed size (Table 1); we previously investigated their seedling responses to microbes extracted from soil cultured by conspecific but not heterospecific adults and seedlings (McCarthy-Neumann & Kobe 2008).
Table 1. Local community characteristics of the six study species used to assess conspecific versus heterospecific plant–soil feedbacks on tropical tree seedling performance (condensed from Table 1, McCarthy-Neumann & Kobe 2008)
Adult basal area3,6
1Species are sorted by seedling shade-tolerance classification and then adult abundance. 2Percent mortality of seedlings at 1% full sun and with zero conspecific seedling density; shade tolerant <10.5%, shade intermediate 10.5–20% and shade intolerant >20%. 3Data presented are means and SDs from mapped stands. 4Number of individuals (≥5 cm d.b.h.) ha−1; rare <3, intermediate 3–10, common >10. 5Mean standing seedling density (≤5years old) ha−1 year−1; rare <20, intermediate 20–200, common >200. 6Total mass (m2 ha−1); rare <0.06, intermediate 0.06–0.25, common >0.25. 7Dry seed mass (mg); small <0.03, medium 0.03–0.30, large >0.30.
This research was conducted at La Selva Biological Station (Sarapiquí Region, Costa Rica) which is operated by the Organization for Tropical Studies. La Selva is a 1510-ha reserve of diverse (> 400 tree species), wet tropical forest receiving c. 4000 mm of rain annually with a mean annual temperature of 25.8 °C (Hartshorn & Hammell 1994).
Soil culturing by conspecific versus heterospecific trees (H1) and conspecific versus heterospecific seedling response to a given extract (H2)
To test the effect of soils cultured by conspecific versus heterospecific trees on seedling performance and to test the relative success of conspecific versus heterospecific seedlings in a given soil extract, we collected soil cores beneath adults of each of the study species. We removed a 10-cm diameter by 30-cm deep soil core within 1 m from the bole of four adults for each species in December 2005. Sampled trees were randomly selected from adults with a diameter at breast height (d.b.h.) at ≥75th percentile for that species located in five 41 × 241 m mapped stands. To minimize potential for multi-species culturing of soil, we randomly selected adults for which no individuals >5, >9 or >20 cm d.b.h. of the other five species were located within 5, 10 or 20 m, respectively, from the focal tree. To further culture the soils, each core was planted with four germinating conspecific seeds, a density that was maintained for 13 weeks in a shade house at 1% full sun. Seeds were collected within 5 m from the trail system throughout La Selva, surface-sterilized (0.6% NaOCl solution) and weighed prior to planting.
Sterilization effect on pathogen infection (H3)
Before obtaining microbial extracts, roots of the seedlings that were used to culture the soil were cut and mixed with the soil. Above-ground seedling portions were discarded. Microorganisms <20 μm were extracted from cultured soil using a wet-sieving method adapted from Klironomos (2002). For each extraction, 40 g of soil was blended with 250 mL of water for 30 s. The liquid suspension was washed through 250, 45 and 20-μm analytical sieves with tap water, keeping the extract to ≤800 mL. Sieves were ultrasonically cleaned for 5 min between each extraction, which minimized, if not eliminated, contamination between treatments. To test for microbial effects on seedling performance (H3), planted seedling pots were amended with autoclaved (30 min at 121 °C) versus unsterilized soil extract. Sterile soil extracts taken near individual adult trees of the same species were combined, resulting in 36 treatments (6 species of seedling × 6 species of adult culturing); for the non-sterile soil extracts, all culturing treatments (e.g. 6 species of seedling × 6 species of adult culturing × 4 trees per species of adult culturing) were kept separate throughout the experiment.
In a previous shadehouse experiment under similar conditions (S. McCarthy-Neumann & R.K. Kobe, unpublished data), arbuscular mycorrhizas (AMF) enhanced seedling mortality, which may have arisen from the additional carbon cost of establishing a mycorrhizal network in the shadehouse pots compared with field conditions. Therefore, we used microbial extracts instead of whole soil to assess plant–soil feedbacks. This was possible because we were able to isolate the pathogen and saprobic fraction (filtrate <20 μm) for the treatments while eliminating AMF spores (collected on the 45-μm sieve).
Planting methods and irradiance effect on pathogen infection (H4)
Within 5 days of obtaining extract (which was stored at 4 °C), seeds with newly emerged radicles were planted into pots (7.5 cm diameter × 25 cm depth) filled with a 1:4 mixture of sterilized field soil and commercial peat moss (Nutripeat; Sun Grow Horticulture Canada Ltd, Vancouver, BC, Canada). Field soil was collected from a common pit in a residual, secondary forest at La Selva and was autoclaved for 2 h at 121 °C followed by 2 days incubation and a second autoclaving. Lethal temperatures (≥ 121 °C) were confirmed at the centre of each soil bag. Each experimental seedling received 100 mL of non-sterilized or sterilized extract. To test for irradiance effects (H3), seedlings were grown at 1% and 5% full sun. Light levels were chosen to mimic endpoints typically encountered from understorey to small tree fall gaps at La Selva Biological Station (Chazdon & Fetcher 1984). To achieve light levels, we used an inner layer of black shade cloth with an outer layer of reflective knitted poly-aluminium shade cloth to minimize heat build-up. Light availability as a percentage of full sun levels was confirmed at each bench by calculating the percentage of sunlight measured in the open and with paired photosynthetic active radiation (PAR) measurements at each bench in the shade houses. PAR was measured on a uniformly overcast day with a LI-COR 250A quantum sensor (LI-COR, Lincoln, NE, USA).
Experimental treatments, seedling measurements and harvesting
To summarize, experimental treatments consisted of species of tree culturing the soil, sterilization and irradiance level. We also had an additional control of seedlings grown with only tap water and no soil extract. The 3084 seedlings were randomly assigned to 10 benches (six for low and four for high light) and were allocated among treatments per Appendix S1.
Emergence and survival were censused and seedlings were watered (∼50 mL of deionized water) by hand twice weekly for 12 weeks (March–June 2005). We assigned date of death as the first census where total leaf or stem tissue necrosis was encountered, at which time seedlings were harvested. To isolate pathogens, roots were surface-sterilized for 1 min with 0.6% NaOCl and cross-sections from the edge of the disease lesion were plated on water agar amended with Ampicillin. Isolates were identified to genus. To determine mass, we harvested seedlings surviving to the end of the experiment, washed soil from roots and oven-dried living tissue at 70 °C to constant mass.
Statistical analysis of experimental seedlings
Seedling survival times were analysed with survival analysis (spss v. 15.0; SPSS Inc, Chicago, IL, USA) and included both pre- and post above-ground emergence stages. Cox proportional hazards regression (Cox & Oakes 1984) tested relative effects of soil source (species culturing soil), sterilization, light availability and initial seed mass on survival times.
We tested treatment effects (soil source, sterilization and irradiance level) and their interactions on growth responses with split-plot ancova, split for light, using bench as a blocking factor (spss). Because seed mass can influence seedling size, estimated dried embryo mass (based on regressions of dry embryo mass to fresh seed mass developed from 20 to 40 randomly selected seeds for each species) was a covariate. Data were ln-transformed when errors were not normal. We ran full models (main treatments, bench, covariate and interactions) for each dependent variable and species to test the assumption that covariate effects were independent of treatment effects; interaction terms were removed when P >0.05. If either terms for bench, covariate or interaction between main treatments had P >0.25 (Bancroft 1964), then the highest order term with the highest P value was removed, and the analysis was run with a reduced model. This process was repeated until all terms with P >0.25 were removed. Adjusted means were compared when the covariate was retained and raw means when the model was reduced to anova. When main effect of soil source was significant (P <0.10), a Holm adjustment was used to compare the conspecific with each of the five heterospecific soil sources for each species. Since the primary interest of this study was to investigate plant–soil feedbacks as they relate to the J–C Model, we did not test for differences among heterospecific soil sources for effects on seedling performance. Differences between treatment means were assessed and P values calculated through degree of overlap in 95% confidence intervals (Austin & Hux 2002).
To summarize reciprocal effects, we compared percentage difference in integrated seedling performance [(mean total mass × mean survival time)/(days of experiment)] for each species pair in soil extract (combined sterile and non-sterile treatments) relative to tap water, averaged across irradiance levels. Our metric of seedling performance encapsulates the race among species from seedling to canopy tree in a space that was previously occupied and cultured by another canopy tree. The winning species for a given set of conditions will attain the largest size among the species that have at least one surviving individual (Kobe 1999). In addition, to determine relative effects of microbial versus chemical factors, we calculated percentage difference in integrated seedling performance for each species pair as: species culturing through chemical effects alone [=seedling performance in sterile extract/seedling performance in tap water] and species culturing through soil microorganisms alone [=(seedling performance in non-sterile extract − seedling performance in sterile extract)/seedling performance in tap water]. We estimated 95% confidence intervals for integrated seedling performance metrics by bootstrapping 3000 data sets for each species (sampling with replacement) and analysing each data set as described above using R (R Development Core Team 2008).
Conspecific culturing had greater influence on seedling performance than heterospecific culturing (H1)
Tree species culturing soil (i.e. soil source) affected survivorship in four (Fig. 1a–d and Table 2; Appendix S2) and total mass in five of six species (Fig. 2 and Table 2; Appendix S3). Seedlings of our study species were more likely to respond negatively in survival and/or growth to soil extract cultured by conspecific than heterospecific individuals.
Table 2. List of significant responses: (a) of seedlings to conspecific versus heterospecific cultured soil extracts and (b) among conspecific versus heterospecific seedlings to soil extract cultured by a given species. Significance shown as follows: *P ≤ 0.10; **P ≤ 0.05; ***P ≤ 0.01; ****P ≤ 0.001
Integrated seedling performance‡
Data condensed from Appendices S2 and S3 and Table 3. P values for multiple comparisons have been adjusted with a Holm adjustment.
†Hazard ratios > 1 indicate reduction in days to mortality of study species.
‡Total mass and integrated seedling performance values are the negative (−) or positive (+) percent difference in seedling response between (a) conspecific versus heterospecific tree soil sources and (b) conspecific versus heterospecific seedlings.
(a) Species’ seedling response to conspecific versus heterospecific cultured soil extracts (H1)
Am vs. Vk (1.59**)
Cs vs. Am (0.3***)
Cs vs. Id in high light (+65%****)
Cs vs. Id in high light (+40%***)
Cs vs. Pm (0.50**)
Cs. vs. Pd (0.56**)
Cs. vs. Vk (0.36***)
Id vs. Cs (2.52****)
Id vs. Pm (−45%****)
Id vs. Cs (+31%***)
Id vs. Pm (0.13****)
Id vs. Pd (−17%*)
Id vs. Pm (−69%****)
Id vs. Pd (0.27****)
Id vs. Pd (−38%****)
Pm vs. Am (−14%**)
Pm vs. Am (−13%**)
Pm vs. Cs (−14%**)
Pm vs. Cs (−11%*)
Pm vs. Pd (−11%*)
Pm vs. Pd (−9%**)
Pm vs. Vk (−12%*)
Pm vs. Vk (−11%*)
Pd vs. Vk (0.26***)
In non-sterile extract and high light
In non-sterile extract and high light
Pd vs. Id (−46%***)
Pd vs. Am (−33%*)
Pd vs. Vk (−37%*)
Pd vs. Cs (−30%***)
Pd vs. Id (−54%****)
Pd vs. Pm (−31%**)
Pd vs. Vk (−42%***)
In non-sterile extract
Vk vs. Am (−25%****)
Vk vs. Am (−25%***)
Vk vs. Cs (−31%****)
Vk vs. Cs (−20%***)
Vk vs. Id in non-sterile extract (−34%****)
Vk vs. Id (−27%****)
Vk vs. Pm (−24%***)
Vk vs. Pm (−10%**)
Vk vs. Pd in non-sterile extract (−24%***)
Vk vs. Pd (−18%***)
Integrated seedling performance‡
(b) Relative response of conspecific versus heterospecific seedlings to soil extract cultured by a given species (H2)
Am vs. Cs (−40%****)
Am vs. Pm (−32%****)
Am vs. Pd (−41%***)
Am vs. Vk (−60%****)
In non-sterile extract
Cs vs. Am (+28%**)
Cs vs. Id (+41%****)
Cs vs. Pm (−5%****)
Cs vs. Pd (−14%****)
Cs vs. Vk (−35%****)
Id vs. Am (+18%**)
Id vs. Cs (+20%***)
Id vs. Pd (−46%****)
Id vs. Vk (−52%****)
Pm vs. Am (+15%**)
Pm vs. Id (−65%****)
Pm vs. Vk (−31%****)
In non-sterile extract and high light
Pd vs. Cs (−24%****)
Pd vs. Id (−45%****)
Pd vs. Pm (−20%**)
Pd vs. Vk (−50%****)
In non-sterile extract
Vk vs. Am (+42%****)
Vk vs. Id (+21%**)
Seedlings of both C. spinosa and P. decurrens had shorter survival time with conspecific than heterospecific soil extract. Results for I. deltoidea were mixed with survival time, either longer or shorter in conspecific than heterospecific cultured soil extract, depending upon the particular species of heterospecific. In contrast, survival time for A. membranacea seedlings was longer with soil extract cultured by conspecific than heterospecific individuals. Survival time for P. macroloba and V. koshnyi seedlings did not vary based on species culturing the soil (Table 2).
Total mass for C. spinosa seedlings was greater with soil extract cultured by conspecific than I. deltoidea individuals but only in high light. In contrast, seedlings of I. deltoidea, P. macroloba and V. koschnyi all had lower mass with soil extract cultured by conspecific than, at least some, if not all, heterospecific individuals (Fig. 2 and Table 2). However, the difference between V. koshnyi seedling mass in extracts cultured by conspecific versus I. deltoidea and P. decurrens individuals only occurred in non-sterile extract. More than 10% of P. decurrens seedlings lost cotyledons prior to harvest. When analysing all seedlings but excluding cotyledon mass, P. decurrens total mass was lower in conspecific than I. deltoidea- and V. koschnyi-cultured non-sterile soil extract in high light. However, when analysing only seedlings that retained cotyledons, total mass for P. decurrens was greater in conspecific than C. spinosa and I. deltoidea soil sources in low light (Table 4.1 and Appendix U, McCarthy-Neumann 2008). These apparently conflicting results are because of greater retention of cotyledon mass in P. decurrens-cultured soil extract (results not shown). The inclusion or exclusion of cotyledon mass did not change results for any other species.
Comparing conspecific and heterospecific culturing on the basis of integrated seedling performance yielded similar results to those obtained from seedling survival and total mass separately (Table 2 and 3). Specifically, conspecific soils were never beneficial and were more likely to be detrimental than at least four heterospecific soils for P. macroloba, P. decurrens and V. koschnyi seedlings. Relative to conspecific soils, I. deltoidea seedlings performed better with culturing by P. macroloba and P. decurrens, worse with culturing by C. spinosa and performed equally well with A. membranacea- and V. koschnyi-culturing. Colubrina spinosa seedlings performed worse with conspecific than I. deltoidea-culturing, but heterospecific culturing by the other four species did not differ from conspecific culturing. Only A. membranacea’s seedling performance was not significantly affected by extract source. Thus, taking the conventional approach of comparing performance in conspecific versus heterospecific soils, conspecific soils were more likely to be detrimental than heterospecific soils (Fig. 3).
Table 3. Reciprocal effects (percent difference in integrated seedling performance [(mean total mass × mean life span)/(days of experiment)]) of plant–soil feedbacks for each study species integrated across extract treatments and irradiance level
Species of adult culturing the soil
Results for Colubrina spinosa feedbacks are from high light, Virola koschnyi feedbacks from non-sterile extract treatment, and Prestoea decurrens feedbacks are from non-sterile extract treatment in high light. Effect of conspecific-cultured soil on seedlings is in bold. Bootstrap-derived 95% CI are in parentheses.
Conspecific seedling performance tended to be disadvantaged relative to heterospecific seedlings in soil extract cultured by a particular tree species (H2)
From the perspective of seedlings contesting space occupied and cultured by a particular canopy tree, conspecific versus heterospecific seedlings were disadvantaged in 15 and favoured in 7 out of 30 cases (Fig. 4 and Table 2 and 3). Soils cultured by A. membranacea and P. decurrens resulted in lower conspecific seedling performance than heterospecific seedlings. Three species (C. spinosa, I. deltoidea and P. macroloba) had mixed results. For example, soil cultured by C. spinosa resulted in worse performance for conspecific seedlings than P. macroloba, P. decurrens and V. koschnyi seedlings, but better performance for conspecific seedlings than A. membranacea and I. deltoidea seedlings. Finally, soils cultured by V. koschnyi individuals resulted in better performance for conspecific seedlings than A. membranacea and I. deltoidea seedlings.
Sterilization minimally influenced seedling performance for one species (H3) and irradiance did not interact with pathogen infection (H4)
Sterilization of soil extracts did not influence survival time or growth in most species, with the exception of P. macroloba, where the non-sterile extract decreased survival time (Hazard ratio = 2.78, P =0.004; Appendix S2) but increased total mass of surviving seedlings relative to sterilized extract across all soil sources (Fig. 2g, h; Appendix S3). The majority of dead seedlings that were identified as diseased were infected by Fusarium (36%), Rhizoctonia (29%) and Phoma (22%).
Light availability did not influence pathogen effects (i.e. no significant light × sterilization interactions) on survival time or growth for any species. Under higher light, survival time was longer for all species except I. deltoidea and V. koschnyi, which were the most shade tolerant of the species (Fig. 5; Appendix S2), and mass was greater for all species except I. deltoidea (Fig. 2; Appendix S3).
In general, chemical factors in soil extracts influenced seedling performance more than biotic factors (Fig. 6). Chemical effects on performance were detrimental for A. membranacea (for all soil sources), C. spinosa (I. deltoidea soil source) and I. deltoidea (for all soil sources except P. macroloba and P. decurrens) seedlings. Chemical effects benefited seedling performance for I. deltoidea (for P. macroloba and P. decurrens soil source), P. decurrens (C. spinosa and V. koschnyi soil sources) and V. koschnyi (A. membranacea, C. spinosa and P. macroloba soil sources). Microbial effects were detrimental for P. decurrens in soil extracts cultured by C. spinosa and increased performance for I. deltoidea seedlings in soil extract cultured by P. macroloba and for P. decurrens seedlings in soil extract cultured by I. deltoidea individuals.
Seedling performance for A. membranacea, I. deltoidea and P. macroloba are based on effects of soil source regardless of extract treatment or light. For the other species, either extract or irradiance treatments interacted with soil source effects. Thus, differences in performance among the different sources of soil extracts are restricted to high light for C. spinosa, non-sterile extract treatment for V. koschnyi and non-sterile extract treatment in high light for P. decurrens seedlings (based on mean mass excluding cotyledon mass).
Our experiment revealed complex interactions among tree species, where the dominant mode of soil feedbacks was mediated through chemical rather than biotic influences and where negative feedbacks from conspecific individuals were stronger than all heterospecific feedbacks for three of the six species. In addition, most heterospecific seedlings had better performance than conspecific seedlings in soil extracts influenced by a given species. However, we only investigated the species-specific effects of plant–soil feedbacks for six species; if we had added more species, we may have found other species with greater negative culturing influence, usurping the primacy of conspecific effects on the three more common species.
Conspecific culturing had larger influence on seedling performance than heterospecific culturing (H1)
Seedling performance was lowest for 16 and highest for only 2 out of 30 cases in soil extract cultured by conspecific versus heterospecific individuals. For three species (P. macroloba, P. decurrens and V. koschnyi), seedling performance was lowest when seedlings were grown with extract cultured by conspecific than all heterospecific individuals. None of the soil sources here were uniformly beneficial or detrimental to seedlings of all species. Likewise, the relative influences of biotic and chemical factors from a given soil extract source could change depending upon the species of responding seedlings; however, chemical effects in soil extracts generally influenced seedling performance much more than biotic effects. As a whole, these results support fairly specific interactions between pairs of species that are mediated mainly through chemical factors with conspecific culturing tending to be more detrimental than heterospecific soil culturing.
An assumption of the J–C Model is that maintaining species diversity requires that NCDD processes are more prevalent in species that are common versus those that are rare, thereby constraining the abundance of common species (Connell, Tracey & Webb 1984). In this study, species that are common as adults from the site where soil was collected (e.g. P. macroloba, P. decurrens and V. koschnyi) tended to perform worse in conspecific versus heterospecific cultured soil extracts. Thus, plant–soil feedbacks could contribute to the coexistence of these tree species even though chemical factors are the main mechanism for reduced performance in soil extracts and not host-specific natural enemies as assumed by the J–C Model.
Conspecific seedling performance was more likely to be reduced than heterospecific seedlings in soil extract cultured by a particular tree species (H2)
Numerous studies similar to this one support the J–C prediction that a focal species’ seedling performance is reduced at near versus far distances from conspecific adults (Augspurger 1983, 1984; Augspurger & Kelly 1984; Packer & Clay 2000; Hood, Swaine & Mason 2004; Bell, Freckleton & Lewis 2006). In contrast, there have been no studies in tropical forests that have tested whether a given species, whose seedlings are disadvantaged in the presence of conspecific adults, would more probably have its adults replaced by heterospecific than conspecific seedlings, leading to greater local plant diversity than expected in the absence of NCDD processes. In our study, heterospecific seedlings were more likely to have better performance than conspecific seedlings in soil extracts influenced by a given individual. However, for many species, the results were mixed so that some heterospecific seedlings were favoured, whereas others were disadvantaged relative to the conspecific seedling. The lack of a consistent disadvantage to conspecific versus heterospecific seedling recruitment may place some limits on the extent that chemically mediated plant–soil feedbacks can enhance tree species coexistence in this wet tropical forest. Thus, knowledge of the relative benefits or costs of ‘escape’ from conspecific adults on seedling recruitment (i.e. H1) does not translate into the ability to predict which species’ seedlings are favoured for recruitment near a particular tree species (i.e. H2).
Our soil source results would not have been interpretable or significant for half of our study species if we had lumped all soil extracts cultured by heterospecific individuals into one treatment (e.g. ‘far’ treatments in Augspurger 1984; Packer & Clay 2000, 2003; Hood, Swaine & Mason 2004). For instance, A. membranacea survival time and C. spinosa mass do not appear to be influenced by soil extract source when comparing conspecific with a lumped heterospecific grouping (results not shown). However, soil extract source does influence seedling performance for these species, but the effect varies among the heterospecific species in magnitude and depends on whether the effect is beneficial or detrimental relative to conspecific derived soil extract (Figs 1 and 2c, d).
Extract culturing by trees of different species may be confounded with an underlying template of soil chemistry that determines tree species occurrence. However, if a mature tree’s long-term survival is because of the underlying micro-conditions, then it is most probably that this soil would be most appropriate for its seedlings too. Our results, however, suggest that seedlings are more likely to be negatively affected by soils from the same microenvironment than by those occupied by a conspecific adult.
Sterilization minimally influences seedling performance for one species (H3)
Surprisingly, the effect of soil extract culturing on seedling performance reflects a chemical rather than a biotic process, since responses to soil culturing were similar in non-sterile and sterile soil extracts. If effects of culturing had been due only to biotic agents, then we would have expected negligible differences among soil origins for sterilized extracts. In a parallel temperate experiment, differences in extract cation availability may have been partially responsible for observed species-specific feedbacks between tree adults and seedlings (McCarthy-Neumann & Kobe 2010). However, there are other possible chemical factor(s) that could mediate plant–soil feedbacks. For instance, tannins in leaf litter can decrease nutrient cycling through decreased decomposition, inhibition of nitrification and other changes in microbial activity (Kraus, Dahlgren & Zasoski 2003). Additionally, leaf litter leachates of some species can suppress seedling performance of other species’ seedlings (Hane et al. 2003; Stinson et al. 2006). Our results provide strong motivation for investigating the chemical factors that could vary among soils cultured by different species of mature trees and their potential impact on seedling dynamics.
Contrary to expectation, negative effects of soil biota were not host-specific and were minimal. Diseased dead seedlings for all species were infected by a variety of soil fungal pathogens (Fusarium, Rhizoctonia and Phoma). However, only one species, P. macroloba, responded to the non-sterile extract with a decrease in survival time (Hazard ratio = 2.78, P =0.004) and a slight increase in mass (Fig. 2g, h), but responses were similar regardless of which adult tree species had cultured the soil. The minimal effect of the soil biota in the extracts was surprising, because in a prior study, seedling performance for these species differed significantly between sterile and non-sterile extract treatments at low light (McCarthy-Neumann & Kobe 2008). In particular, the non-sterile extract from conspecific soil reduced either survival time or total mass for three of the species represented here (A. membranacea, P. macroloba and P. decurrens) and increased survival time for I. deltoidea. We obtain consistent results between these two studies when we restrict our current data set to compare survival time and total mass only for conspecific cultured extract at low light. For instance, survival time is reduced in the non-sterile versus sterile extract treatment for P. macroloba (3%, χ2 = 3.94, d.f. = 1,48, P <0.05) and mass is reduced for A. membranacea (25%, F1,14 = 2.06, P <0.10) and P. decurrens (25%, F1,33 = 3.98, P <0.10) seedlings. The response of seedlings to non-sterile extract cultured by various species was highly diverse (Fig. 6) and may be partially responsible for the lack of a widespread effect of soil microbes on seedling performance in this study. In addition, excluding the filtrate component > 20 μm in soil extracts may have eliminated some pathogenic microbes, underestimating their role.
Irradiance did not interact with pathogen infection (H4)
Increased irradiance did not ameliorate the negative effect of soil microbes on seedling performance as we had predicted. Indeed P. macroloba’s survival time and growth were affected by soil microbes without regard to irradiance level, whereas soil origin differences for P. decurrens seedling growth were only different in the non-sterile treatment at moderately high light. These results are consistent with those from a parallel experiment working with temperate species, but are inconsistent with studies using different methodologies that would allow abiotic factors (e.g. soil moisture and temperature) and/or AMF colonization to vary between irradiance levels (Augspurger 1984; Augspurger & Kelly 1984; Hood, Swaine & Mason 2004). In addition, light effects on seedling susceptibility to soil microbes might have been detected if our higher light treatment had been > 5% full sun and had reflected large rather than small tree fall gaps.
Plant–soil interactions affected seedling performance by as much as 60%, which is comparable in magnitude to the effect (c. 0–130%) that an increase in light from 1% to 5% full sun had on mean seedling mass for the same species. This moderate increase in light encompasses the endpoints typically encountered from understorey to small tree fall gaps at La Selva Biological Station (Chazdon & Fetcher 1984). Thus, plant–soil feedbacks could have a large influence (perhaps similar in magnitude to that of spatial heterogeneity in irradiance) on community dynamics and composition in tropical forests.
Results from this experiment may have been influenced by the soil extraction methodology used to obtain the soil source treatments. For instance, a meta-analysis by Kulmatiski et al. (2008) suggests that experiments using cultivated soils and glasshouse conditions tend to be more detrimental to plant performance than experiments using field-collected soils and field conditions. Conversely, it is possible that the minimal negative effect of soil biota was because of excluding the filtrate component > 20 μm, which may have eliminated some pathogenic microbes. Moreover, since we excluded AMF spores, our results do not necessarily reflect the net effect of the soil microbial community on seedling recruitment patterns in the field. For instance, we may have underestimated the negative effect of soil biota, if mycorrhizal colonization in low-light environments is also detrimental in field conditions.
Comparison between tropical and temperate forests
There is accumulating evidence of NCDD seedling mortality and reduced growth (Packer & Clay 2000; Hille Ris Lambers, Clark & Beckage 2002; Packer & Clay 2003) in temperate systems, which has called into question the long-standing assumption of the J–C Model that NCDD processes are less pronounced in temperate than tropical forests. We explicitly tested this assumption by directly comparing seedling responses to NCDD processes between tropical and temperate species to parallel experiments designed to enable direct comparison. We found that soil-mediated (mostly chemical factors) species-specific feedbacks between tree adults and seedlings were ubiquitous in both temperate and tropical systems. However, the feedbacks between temperate species were largely idiosyncratic and were inconsistent with the J–C Model, since soils cultured by heterospecific species were more likely to decrease seedling performance than soils cultured by conspecific individuals (Fig. 3; McCarthy-Neumann & Kobe 2010). In addition, soils cultured by a particular species did not necessarily improve heterospecific seedling performance relative to conspecific seedlings (Fig. 4; McCarthy-Neumann & Kobe 2010). In contrast, in the tropical forest, species tended to perform worse in conspecific than heterospecific cultured soil extract (Fig. 3), especially species that were common as adults from the site where the soil was collected (Table 1). In addition, soils cultured by a particular species were more likely to favour heterospecific performance relative to conspecific seedling performance (Fig. 4). An important caveat to these results is that NCDD was only investigated for six tropical species; investigating additional species may usurp the primacy of conspecific effects on common tree species. In addition, at this time, we do not know the mechanism for why common species may be influenced more negatively by chemical factors in conspecific cultured soils than less common species. However, our results do suggest that chemical mediated plant–soil feedbacks are an important component of seedling dynamics in both temperate and tropical forests, that these feedbacks can create complex interactions between tree species and that currently there is more evidence that these feedbacks may enhance species coexistence in tropical than temperate forests.
We thank Mindy McDermott, Marisol Luna, Ademar Hurtado, Martin Cascante, Edgar Vargas and David Neumann for their invaluable assistance. We also thank Michael Walters, John Klironomos, Andrew Jarosz, David Rothstein, Tom Baribault, Ellen Holste, Corine Vriesendorp and two anonymous referees for comments on earlier drafts of this manuscript and Andrew Finley for statistical help. The Organization for Tropical Studies and the staff at La Selva Biological Station provided logistic support. This research was supported by the National Science Foundation (DEB 0235907, 0075472, 0743609).