Landslides significantly alter land cover and the distribution of biomass: an example from the Ninole ridges of Hawai'i

In the Ninole ridges of Hawai'i we investigated how landslides influence ecosystem development and modify land cover and the distribution of biomass. We estimated above and below-ground biomass, and N and P concentration in leaves (Metrosideros polymorpha) and very fine roots (all species), for vegetation developing on landslides of three age classes (young, < 18 yr; intermediate, 42 yr; and old ca. 124 yr) and on undisturbed soils (ca. 430 yr). The undisturbed soils were derived from ash underlain by basalt. To quantify changes in land cover and the distribution of biomass we combined our estimates of biomass with estimates of the area covered by each vegetation class. The latter estimates were obtained from the analysis and classification of color-infrared aerial photographs. Average above- and below-ground biomass for the herbaceous vegetation (young landslides) was 10.4 and 3.2 t/ha, whereas for the ohia-non ash forest (intermediate and old landslides) was 37.5 and 5.2 t/ha, respectively. For the ohia-ash forest (undisturbed sites), average above and below-ground biomass was 354.6 and 9.5 t/ha, respectively. Average foliar N for the herbaceous and ohia-non ash forest ranged between 0.80–0.84%, whereas root P between 0.056–0.040%, respectively. For the ohia-ash forest, average foliar and root P was 0.918% and 0.036%, respectively. Based on changes in vegetation cover during the last 430 yr, we estimated rate of disturbance at 15% per century or equivalently that 53 t/ha biomass per century exited through the system. The removal of ash-derived soils by landslides significantly alters successional trajectories and by doing so may be transforming the Ninole ecosystems in irreversible ways.     

Heterospecific alarm call recognition in a non-vocal reptile

The ability to recognize and respond to the alarm calls of heterospecifics has previously been described only in species with vocal communication. Here we provide evidence that a non-vocal reptile, the Galápagos marine iguana (Amblyrhynchus cristatus), can eavesdrop on the alarm call of the Galápagos mockingbird (Nesomimus parvulus) and respond with anti-predator behaviour. Eavesdropping on complex heterospecific communications demonstrates a remarkable degree of auditory discrimination in a non-vocal species.     

Uplift,Erosion, and Phosphorus Limitation in Terrestrial Ecosystems

ABSTRACT Primary productivity on old, weathered soils often is assumed to be limited by phosphorus (P), especially in the lowland tropics where climatic conditions promote the rapid depletion of rock-derived nutrients. This assumption is based on a static view of soils weathering in place with no renewal of the bedrock source. In reality, advection of material through the soil column introduces a spatially variable supply of rock-derived nutrients. This flux is dependent on the residence time of soil, which can range from a few hundred years in rapidly uplifting collisional mountain belts to tens of millions of years in tectonically quiescent tropical cratons. We modeled the effects of tectonic uplift, erosion, and soil depth on the advection of P through the soil column and P availability, calibrating rate of change in biologically available P over time with data from two basaltic chronosequences in Hawai’i and a series of greywacke terraces in New Zealand. Combining our model with the global distribution of tectonic uplift rates and soil depths, we identified tectonic settings that are likely to support P-depleted ecosystems—assuming that tectonic uplift and erosion are balanced (that is, landscape development has reached steady state). The model captures the occurrence of transient P limitation in rapidly uplifting young ecosystems where mineral weathering is outpaced by physical erosion—a likely occurrence where biological N fixation is important. However, we calculate that P depletion is unlikely in areas of moderate uplift, such as most of Central America and Southeast Asia, due to the continuous advection of P into the rooting zone. Finally, where soil advection is slow, such as the Amazon Basin, we expect widespread P depletion in the absence of exogenous nutrient inputs.     

Interactive effects of elevated CO<Subscript>2</Subscript>, N deposition and climate change on plant litter quality in a California annual grassland

Although global changes can alter ecosystem nutrient dynamics indirectly as a result of their effects on plant litter quality, the interactive effects of global changes on plant litter remain largely unexplored in natural communities. We investigated the effects of elevated CO₂, N deposition, warming and increased precipitation on the composition of organic compounds in plant litter in a fully-factorial experiment conducted in a California annual grassland. While lignin increased within functional groups under elevated CO₂, this effect was attenuated by warming in grasses and by water additions in forbs. CO₂-induced increases in lignin within functional groups also were counteracted by an increase in the relative biomass of forbs, which contained less lignin than grasses. Consequently, there was no net change in the overall lignin content of senesced tissue at the plot level under elevated CO₂. Nitrate additions increased N in both grass and forb litter, although this effect was attenuated by water additions. Relative to changes in N within functional groups, changes in functional group dominance had a minor effect on overall litter N at the plot level. Nitrate additions had the strongest effect on decomposition, increasing lignin losses from Avena litter and interacting with water additions to increase decomposition of litter of other grasses. Increases in lignin that resulted from elevated CO₂ had no effect on decomposition but elevated CO₂ increased N losses from Avena litter. Overall, the interactions among elements of global change were as important as single-factor effects in influencing plant litter chemistry. However, with the exception of variation in N, litter quality had little influence on decomposition over the short term.     

Rapid nutrient cycling in leaf litter from invasive plants in Hawai’i

Physiological traits that contribute to the establishment and spread of invasive plant species could also have impacts on ecosystem processes. The traits prevalent in many invasive plants, such as high specific leaf areas, rapid growth rates, and elevated leaf nutrient concentrations, improve litter quality and should increase rates of decomposition and nutrient cycling. To test for these ecosystem impacts, we measured initial leaf litter properties, decomposition rates, and nutrient dynamics in 11 understory plants from the Hawaiian islands in control and nitrogen + phosphorus fertilized plots. These included five common native species, four of which were ferns, and six aggressive invasive species, including five angiosperms and one fern. We found a 50-fold variation in leaf litter decay rates, with natives decaying at rates of 0.2–2.3 year^–1 and invaders at 1.4–9.3 year^–1. This difference was driven by very low decomposition rates in native fern litter. Fertilization significantly increased the decay rates of leaf litter from two native and two invasive species. Most invasive litter types lost nitrogen and phosphorus more rapidly and in larger quantities than comparable native litter types. All litter types except three native ferns lost nitrogen after 100 days of decomposition, and all litter types except the most recalcitrant native ferns lost >50% of initial phosphorus by the end of the experiment (204–735 days). If invasive understory plants displace native species, nutrient cycling rates could increase dramatically due to rapid decomposition and nutrient release from invasive litter. Such changes are likely to cause a positive feedback to invasion in Hawaii because many invasive plants thrive on nutrient-rich soils.     

Analysis of factors regulating ecosystemdevelopment on Mauna Loa using the Century model

We used the Century model to evaluateenvironmental controls over ecosystem developmentduring the first 3500 y of primary succession onpahoehoe (i.e., relatively smooth, solid) lava flowsof wet, windward Mauna Loa, Hawaii. The Century modelis a generalized ecosystem model that simulatescarbon, nitrogen and phosphorus dynamics forplant-soil systems. Preliminary results indicated theneed to modify the model to include the effects ofsoil C accumulation on soil water storage anddrainage. The modified model was parameterized tosimulate observed values of aboveground productivity,biomass and soil element pools on a 3400-y-old site at700 m elevation. Testing the model parameters at 1660m elevation indicated that N inputs were lower andsoil water drainage rates were slower at the higherelevation. We applied the modified and fullyparameterized model to simulate ecosystem attributesduring primary succession at five elevations, andconducted single-factor experiments with the model toidentify the specific influences of variations intemperature, nutrient inputs, and rainfall on modeledecosystem characteristics.Simulated aboveground productivity (ANPP), net N andP mineralization, and biomass element pools allincreased through time at each elevation, and alldeclined with increasing elevation at each point intime. After 3500 y of succession none of theseattributes had reached a stable asymptote, butasymptotes were approached more quickly, andsuccession was therefore faster, at lower than athigher elevations. Simulated soil organic matter(SOM) pools increased with elevation, despite thatplant productivity declined. These results, andsimilar comparisons among rainfall regimes, suggestthat SOM pools were more sensitive to factorscontrolling decay than production rates.Within elevations and temperature regimes, nutrientavailability was the most important factor controllingsimulated rates of plant productivity, biomass, anddetritus accumulation during ecosystem development. Through time, SOM accumulations alleviated nutrientlimitations to plants, but simulated productivityremained highly dependent upon externally suppliednutrients even after 20,000 y. Rainfall had two maineffects on nutrient availability within the model: (1)it increased rates of leaching, and thus depletednutrient supplies; and (2) it exacerbated soil floodingand thereby decreased nutrient turnover rates. Highrainfall on windward Mauna Loa maintains oligotrophicconditions through time despite continuous N and Pinputs.     

Regulation of soil phosphatase and chitinase activityby N and P availability

Soil microorganisms and plants produce enzymes thatmineralize organically bound nutrients. When nutrientavailability is low, the biota may be able to increase production ofthese enzymes to enhance the supply of inorganicnitrogen (N) and phosphorus (P). Regulation of enzyme productionmay be a point where N and P cyclesinteract. We measured acid phosphatase and chitinase(N-acetyl ß-D-glucosaminide) activity in soilacross a chronosequence in Hawaii where N and Pavailability varies substantially among sites and longterm fertilizer plots had been maintained for over 4years.Phosphatase activity was high at all sites. Chitinaseactivity decreased significantly as age and Navailability increased across the chronosequence.Phosphorus addition suppressed phosphatase activity atall sites, while N addition increased phosphataseactivity at the young, N-limited site. In contrast,N addition repressed chitinase activity only at the Nlimited young site, and P additions had no effect onchitinase activity. These results suggest that theregulatory relationship between nutrient supply andnutrient mineralization are asymmetric for N and P,and that the differences could help to explaindifferences observed in patterns of N and Pavailability.     

Changes in Asymbiotic, Heterotrophic Nitrogen Fixation on Leaf Litter of Metrosideros polymorpha with Long-Term Ecosystem Development in Hawaii

We measured nitrogenase activity (acetylene reduction) of asymbiotic, heterotrophic, nitrogen-fixing bacteria on leaf litter from the tree Metrosideros polymorpha collected from six sites on the Hawaiian archipelago. At all sites M. polymorpha was the dominant tree, and its litter was the most abundant on the forest floor. The sites spanned a soil chronosequence of 300 to 4.1 million y. We estimated potential nitrogen fixation associated with this leaf litter to be highest at the youngest site (1.25 kg ha^-1 y^-1), declining to between 0.05 and 0.22 kg ha^-1 y^-1 at the oldest four sites on the chronosequence. To investigate how the availability of weathered elements influences N fixation rates at different stages of soil development, we sampled M. polymorpha leaf litter from complete, factorial fertilization experiments located at the 300-y, 20,000-y and 4.1 million–y sites. At the youngest and oldest sites, nitrogenase activity on leaf litter increased significantly in the plots fertilized with phosphorus and “total” (all nutrients except N and P); no significant increases in nitrogenase activity were measured in leaf litter from treatments at the middle-aged site. The results suggest that the highest rates of N fixation are sustained during the “building” or early phase of ecosystem development when N is accumulating and inputs of geologically cycled (lithophilic) nutrients from weathering are substantial. Received 4 February 1999; accepted 29 March 2000.     

Are patterns in nutrient limitation belowground consistent with those aboveground: results from a 4 million year chronosequence

Accurately predicting the effects of global change on net carbon (C) exchange between terrestrial ecosystems and the atmosphere requires a more complete understanding of how nutrient availability regulates both plant growth and heterotrophic soil respiration. Models of soil development suggest that the nature of nutrient limitation changes over the course of ecosystem development, transitioning from nitrogen (N) limitation in ‘young’ sites to phosphorus (P) limitation in ‘old’ sites. However, previous research has focused primarily on plant responses to added nutrients, and the applicability of nutrient limitation-soil development models to belowground processes has not been thoroughly investigated. Here, we assessed the effects of nutrients on soil C cycling in three different forests that occupy a 4 million year substrate age chronosequence where tree growth is N limited at the youngest site, co-limited by N and P at the intermediate-aged site, and P limited at the oldest site. Our goal was to use short-term laboratory soil C manipulations (using ¹⁴C-labeled substrates) and longer-term intact soil core incubations to compare belowground responses to fertilization with aboveground patterns. When nutrients were applied with labile C (sucrose), patterns of microbial nutrient limitation were similar to plant patterns: microbial activity was limited more by N than by P in the young site, and P was more limiting than N in the old site. However, in the absence of C additions, increased respiration of native soil organic matter only occurred with simultaneous additions of N and P. Taken together, these data suggest that altered nutrient inputs into ecosystems could have dissimilar effects on C cycling above- and belowground, that nutrients may differentially affect of the fate of different soil C pools, and that future changes to the net C balance of terrestrial ecosystems will be partially regulated by soil nutrient status.     

Erosion, Geological History, and Indigenous Agriculture: A Tale of Two Valleys

Irrigated pondfields and rainfed field systems represented alternative pathways of agricultural intensification that were unevenly distributed across the Hawaiian Archipelago prior to European contact, with pondfields on wetter soils and older islands and rainfed systems on fertile, moderate-rainfall upland sites on younger islands. The spatial separation of these systems is thought to have contributed to the dynamics of social and political organization in pre-contact Hawai’i. However, deep stream valleys on older Hawaiian Islands often retain the remains of rainfed dryland agriculture on their lower slopes. We evaluated why rainfed agriculture developed on valley slopes on older but not younger islands by comparing soils of Pololū Valley on the young island of Hawai’i with those of Hālawa Valley on the older island of Moloka’i. Alluvial valley-bottom and colluvial slope soils of both valleys are enriched 4–5-fold in base saturation and in P that can be weathered, and greater than 10-fold in resin-extractable P and weatherable Ca, compared to soils of their surrounding uplands. However, due to an interaction of volcanically driven subsidence of the young island of Hawai’i with post-glacial sea level rise, the side walls of Pololū Valley plunge directly into a flat valley floor, whereas the alluvial floor of Hālawa Valley is surrounded by a band of fertile colluvial soils where rainfed agricultural features were concentrated. Only 5% of Pololū Valley supports colluvial soils with slopes between 5° and 12° (suitable for rainfed agriculture), whereas 16% of Hālawa Valley does so. The potential for integrated pondfield/rainfed valley systems of the older Hawaiian Islands increased their advantage in productivity and sustainability over the predominantly rainfed systems of the younger islands.