Litter flammability in oak woodlands and shrublands of southeastern France

Characterizing the flammability of litter fuels is of major importance for assessing wildland fire ignition hazard. Here we compared the flammability of litter within a mosaic of Quercus suber (cork oak) woodlands and shrublands in a Mediterranean fire-prone area (Maures massif, southeastern France) to test whether the characteristics and the flammability of litter vary with the vegetation types. We tested experimentally the ignitability, the sustainability, the combustibility and the consumability of undisturbed (=non-reconstructed) litter samples with a point-source mode of ignition. Although the frequency of ignition was similar between all the vegetation types, we distinguished four groups having litter of specific composition and flammability: low and sparse shrublands dominated by Cistus species, medium shrublands with cork oak, high Erica shrublands with sparse cork oak woodlands, and mixed mature oak woodlands with Q. suber, Q. ilex and Q. pubescens. As these vegetation types corresponded to a specific range of past fire recurrence, we also tested the effect of the number of fires and the time since the last fire on litter flammability. Litters of plots recurrently burned had low ability to propagate flames and low flame sustainability. We discuss how the recent fire history can modify vegetation and litter flammability, and thus the fire ignition hazard.     

Fire-driven oak regeneration in French Mediterranean ecosystems

Regeneration by seeds for cork oak (Quercus suber) and companion oaks (holm oak Quercus ilex and downy oak Quercus pubescens) is likely to be poor in the fire-prone Maures massif (southern France) but the causes are poorly known. Our objective was to assess the effective recruitment for these three oak species and their temporal pattern of recruitment, in order to determine the main limitation factors and the regeneration window of each species. We studied oak recruits (height <3 m) in naturally regenerated populations according to a gradient of fire recurrence and in five main vegetation types including shrublands and mixed mature woodlands. Fire recurrence was the main explanatory factor of oak recruitment, either directly or through vegetation type and microsite characteristics. The results indicate nil to low recruitment for holm oak and downy oak in shrublands, especially those recurrently burned and dominated by Cistus species. Cork oak recruited better than the other oaks in medium and high shrublands dominated by Erica arborea. In contrast, recruitment was high for holm and downy oak in mixed oak stands and mixed pine-oak stands that have not burned for decades. Microsite conditions such as coverage by litter and shrubs influenced oak recruitment, whereas landscape configuration and stand basal area had no influence. Our results suggest that strategic shrub-clearing, oak planting and protection of mixed oak woodlands as seed sources would help maintaining oak populations in the woodland–shrubland mosaic.     

Long-term understory recovery after mechanical fuel reduction in Mediterranean cork oak forests

In the Euro-Mediterranean region, mechanical fuel reduction is increasingly used in response to the mounting occurrence of catastrophic wildfires, yet their long-term ecological effects are poorly understood. Although Mediterranean vegetation is resilient to a range of disturbances, it is possible that widespread fuel management at short intervals may threaten forest structural complexity and the persistence of some plant species and functional types, with overall negative consequences for biodiversity. We used a chronosequence approach to infer woody vegetation changes in the first 70 years after understory clearing in upland cork oak (Quercus suber) forests, and to assess how these are affected by treatment frequency. Across the chronosequence there was a shift between plant communities with contrasting composition, structure and functional organization. Understory cover increased quickly after disturbance and a community dominated by pioneer seeder and dry-fruited shrubs (Cistus ladanifer, C. populifolius, Genista triacanthos, and Lavandula stoechas) developed during about 15 years, but this was slowly replaced by a community dominated by resprouters and fleshy-fruited species (Arbutus unedo, Erica arborea) >40 years after disturbance. During the first 15 years there were rapid increases in woody species richness, vertical structural diversity, cover by Q. suber juveniles and saplings, and shrub cover at <1.5 m strata, which levelled off or slightly declined thereafter. In contrast, tree species richness, tree density and density of arboreal A. unedo and E. arborea, vertical structural evenness, and cover at >1.5 m strata increased slowly for >50 years. Treatment frequency showed strongly negative relationships with species richness, structural diversity and evenness, and horizontal and vertical understory cover, particularly that of slowly recovering species. These findings suggest that fuel reduction programs involving widespread and recurrent understory clearing may lead to the elimination at the landscape scale of stands with complex multi-layered understory occupied by large resprouters and fleshy-fruited species, which take a long time to recover after disturbance. Fuel management programs thus need to balance the dual goals of fire hazard reduction and biodiversity conservation, recognizing the value of stands untreated for >50 years to retain ecological heterogeneity in Mediterranean forest landscapes.     

Understory vegetation response after 30 years of interval prescribed burning in two ponderosa pine sites in northern Arizona,USA

Southwestern USA ponderosa pine (Pinus ponderosa C. Lawson var. scopulorum Engelm.) forests evolved with frequent surface fires and have changed dramatically over the last century. Overstory tree density has sharply increased while abundance of understory vegetation has declined primarily due to the near cessation of fires. We examined effects of varying prescribed fire-return intervals (1, 2, 4, 6, 8, and 10 years, plus unburned) on the abundance and composition of understory vegetation in 2007 and 2008 after 30+ years of fall prescribed burning at two ponderosa pine sites. We found that after 30 years, overstory canopy cover remained high, while understory plant canopy cover was low, averaging <12% on all burn intervals. We attributed the weak understory response to a few factors – the most important of which was the high overstory cover at both sites. Graminoid cover and cover of the major grass species, Elymus elymoides (squirreltail), increased on shorter fire-return intervals compared to unburned plots, but only at one site. Community composition differed significantly between shorter fire-return intervals and unburned plots at one site, but not the other. For several response variables, precipitation levels appeared to have a stronger effect than treatments. Our findings suggest that low-severity burn treatments in southwestern ponderosa pine forests, especially those that do not decrease overstory cover, are minimally effective in increasing understory plant cover. Thinning of these dense forests along with prescribed burning is necessary to increase cover of understory vegetation.     

A generic process-based SImulator for meditERRanean landscApes (SIERRA): design and validation exercises

Understanding the interrelationship that exists between landscape patterns and fire history requires a great range of case studies to reduce the effects of substrate and climate. The lack of such data has led to an increasing need for spatially explicit models dealing with vegetation dynamics. The challenge is to find a compromise between process complexity, realism and landscape applications. In this paper, we describe a simulation approach (SIERRA) focussed on the particular case of Mediterranean-type communities subjected to large recurrent fires. Firstly, disturbance response strategies used in “vital attributes models” are used to simulate the influence of fire on vegetation dynamics, focussing in particular on the integration of specific regeneration abilities of Mediterranean species. Next, the model takes a functional approach towards carbon and water budgets to drive competition and simulate the seasonal vegetation water status to estimate fire risk. Spatial processes of seed dispersal, surface water fluxes depending on topographic convergence, and fire spread are used to accurately simulate landscape heterogeneity. The model offers a spatial representation of the annual course of vertical structure of biomass and carbon fluxes coupled with the weekly soil water budget and evapotranspiration rates. Some simulation and validation exercises are presented to illustrate both the functional properties on a Quercus ilex stand, and the fire-prone community dynamics of a maquis shrubland. These initial results will form a strong basis for using the model to test hypotheses about fire-prone landscape patterns.     

Predicting stand damage and tree survival in burned forests in Catalonia (North-East Spain)

The study developed models for predicting the post-fire tree survival in Catalonia. The models are appropriate for forest planning purposes. Two types of models were developed: a stand-level model to predict the degree of damage caused by a forest fire, and tree-level models to predict the probability of a tree to survive a forest fire. The models were based on forest inventory and fire data. The inventory data on forest stands were obtained from the second (1989–1990) and third (2000–2001) Spanish national forest inventories, and the fire data consisted of the perimeters of forest fires larger than 20 ha that occurred in Catalonia between the 2nd and 3rd measurement of the inventory plots. The models were based on easily measurable forest characteristics, and they permit the forest manager to predict the effect of stand structure and species composition on the expected damage. According to the stand level fire damage model, the relative damage decreases when the stand basal area or mean tree diameter increases. Conversely, the relative stand damage increases when there is a large variation in tree size, when the stand is located on a steep slope, and when it is dominated by pine. According to the tree level survival models, trees in stands with a high basal area, a large mean tree size and a small variability in tree diameters have a high survival probability. Large trees in dominant positions have the highest probability of surviving a fire. Another result of the study is the exceptionally good post-fire survival ability of Pinus pinea and Quercus suber.     

Phenology and growth dynamics in Mediterranean evergreen oaks: Effects of environmental conditions and water relations

Budburst date and shoot elongation were measured in two mature Mediterranean evergreen oaks (Quercus suber and Quercus ilex) and their relationships with meteorological and tree water status (predawn leaf water potential) data were analysed. Experimental work took place at two sites: Mitra 2 - Southern Portugal (2002-2003) and Lezirias - Central Portugal (2007-2010). Quercus suber phenology was studied at both sites whereas Q. ilex was only studied at Mitra 2. Quercus suber budburst date occurred at a photoperiod around 13.8 h (± 0.26) - late April/early May - and was highly related to the average daily temperature in the period 25 March - budburst date (ca. 1.5 months prior to budburst), irrespective of site location. In that period, budburst date was much more dependent on average maximum than average minimum daily temperature. Base temperature and thermal time for Q. suber were estimated as 6.2 °C (within the reported literature values) and 323 degree-days, respectively. Q. ilex budburst occurred about 6 weeks earlier than in Q. suber (photoperiod: 12.3 h (±0.3)). Relationships of Q. ilex budburst date and temperature were not studied since only 2 years of data were available for this species. Q. suber shoot elongation underlying mechanisms were quite different in the two sites. At Mitra 2 (Q. suber and Q. ilex), there was a considerable tree water stress during the dry season which restricted shoot elongation. Shoot growth was resumed later in the wet autumn when tree water status recovered again. At the Lezirias site Q. suber water status was not restrictive. Therefore, shoot elongation was mainly dependent on nutrient availability in top soil, as suggested by the strong and positive relationships between annual shoot growth and long-term cumulative rainfall (2-4 months) and short-term average temperature (1 month) prior to budburst. Annual shoot elongation at this well-watered site was higher than in Mitra 2, and variability of growth between trees was enhanced after warm, wet springs when shoot elongation was higher. Results obtained are relevant to the carbon balance, productivity and management of evergreen Mediterranean oak woodlands, particularly under the foreseen climate change scenarios.     

Canopy cover and groundlayer vegetation dynamics in a fire managed eastern sand savanna

Savanna vegetation is characterized by high and variable ground layer species richness regulated by functional group interactions with fire regimes and canopy cover. Frequent fire selects for C4 grasses and prairie forbs in canopy openings and C3 graminoid species and shade-adapted forbs and shrubs in canopy shade. Frequent fire also maximizes heterogeneity in partial canopy cover and species richness across the full canopy gradient. However, few studies have linked fire induced change in tree canopy cover with groundlayer vegetation dynamics in relation to this model. In 1986 and in 2007, we measured canopy cover and sampled groundlayer vegetation in 1 m² plots along 53 transects at the Tefft Savanna, a fire managed 197 ha eastern sand savanna with strong canopy cover and elevation gradients. We analyzed temporal change in canopy cover and groundlayer vegetation, correlating percent change in canopy cover with change in ground layer functional groups. After 20 years of burning at 3 fires/decade, elevation accounted for 62% of the variation in an NMS ordination of groundlayer vegetation. However, canopy cover, which averaged 24-86% in 2007, had a significant secondary effect on the ordination. Five vegetation types classified by TWINSPAN varied significantly in elevation and canopy cover. Woody vegetation comprised 8 of the 12 species with greatest niche breadths, and tended to predominant in woodland or forest, where tree cover averaged 50% or more. Forbs had greater richness in savanna, which averaged less than 30% canopy cover. The C3 sedge Carex pensylvanica was the dominant graminoid species under woodland canopy cover, and was co-dominant with the C4 grasses Andropogon scoparius and Sorghastrum nutans under savanna canopy cover. As in other savannas, N-fixing species sorted across shade and canopy openings, and heterogeneity among transects was maximized at mid-canopy cover. Over time, canopy cover decreased up to 50%, creating more open savanna conditions at mid to high elevations. This decrease was associated with a 20-100 % increase in species richness and was significantly correlated with increasing richness and cover of C4 grasses and summer flowering prairie and woodland forbs. These results support a canopy cover model of fire-maintained savanna vegetation, with greater abundance of C4 grasses and prairie forb species associated with lower canopy cover, greater heterogeneity at mid-canopy cover, and species richness maximized across the light gradient. They also indicate that decreasing canopy cover caused by repeated burning increases species richness and abundance of C4 and prairie forb species.     

Birds as predators of cork and holm oak pests

The recent decline of Mediterranean oak woodlands in SW Iberian Peninsula is related to insect pests which affect both cork oak (Quercus suber) and holm oak (Quercus rotundifolia). We identified twenty-six bird species as potential regular predators of twenty major pests by reviewing the diet of breeding, wintering and resident species in this ecosystem. Foraging guilds are strongly associated with predation at distinct stages of the pests’ life-cycle: ground-foragers prey on overwintering pupae and larvae of seed-borers, tree-foragers prey on eggs, larvae and pupae of defoliating and wood-boring pests, and aerial-sweepers prey on airborne imagines. Bird predation can cover the complete life-cycle of pest species because different species may be complementary due to a dissimilar exploitation of foraging niches and periods. Small generalist tree-foraging passerines are important pest predators given their high densities and widespread distribution in Mediterranean oak woodlands, but management practices can have a significant negative effect in their populations.     

Shrub species affect distinctively the functioning of scattered Quercus ilex trees in Mediterranean open woodlands

A low tree stand density has been showed as necessary to thrive with summer drought in semiarid Mediterranean open woodlands. Shrub encroachment of these open woodlands is currently recommended to guarantee the persistence of the system, due to the nursery effect of shrubs on tree seedling. However, the increase in abundance and cover of a shrub understory in these water limited woodlands could bring consequences to tree overstory functioning. The present study analyzes the physiological status of scattered Quercus ilex L. trees in paired adjacent plots with and without the presence of a shrubby understory in CW Spain. Two contrasting shrub strategies were addressed in order to take into account possible species-specific effects: a dense-shallow rooting shrub (Cistus landanifer L.) and a sparse-deep rooting shrub (Retama sphaerocarpa (L.) Boiss). Leaf water potential (at predawn and midday), leaf gas exchange parameters (net photosynthetic rate and stomatal conductance), leaf nitrogen content and chlorophyll fluorescence transients (maximum photochemical efficiency and performance index, sensuStrasser et al., 2004) were measured during three consecutive summers. Trees growing with Cistus as understory showed significant lower leaf water potential, leaf gas exchange parameters, leaf nitrogen content and chlorophyll photochemical efficiency than trees growing without shrub competence. However, the presence of the legume Retama did not affect significantly the physiological state of Q. ilex. Thus, we conclude that the presence of a shrubby understory has the potential to modify the functioning of scattered trees, but these effects are species-specific.     

Root system of Quercus suber L. seedlings in response to herbaceous competition and different watering and fertilisation regimes

Different management practices are commonly applied to increase pasture yield of Mediterranean open woodlands, but the consequences of increasing competition for soil resources with these practices on tree recruitment are still unknown. In a greenhouse study, belowground competition of Quercus suber L. seedlings growing together with natural (Oak_NP) or improved pasture (Oak_IP) was evaluated, and their root systems compared with those of seedlings growing in bare soil (Oak_BS). Two watering levels and two regimes of P₂O₅ fertilisation were also tested. Because of competition, the Oak_IP seedlings had their fine root mass density, coarse root length, and shoot mass reduced by up to 40, 36, and 39%, respectively, when compared to Oak_NP seedlings. Oak_NP and Oak_BS seedlings showed similar average root density parameters and shoot mass values, indicating that Q. suber seedlings growing with natural pasture is a low competitive system. High availability of water and phosphorus did not mitigate the strength of competition between herbaceous plants and oak seedlings, and favoured the pasture to the detriment of the trees. Our findings suggest that P₂O₅ fertilisation and irrigation practices performed to improve herbaceous productivity will negatively influence recruitment of Q. suber seedlings.     

Eight years of seasonal burning and herbicidal brush control influence sapling longleaf pine growth,understory vegetation,and the outcome of an ensuing wildfire

To study how fire or herbicide use influences longleaf pine (Pinus palustris Mill.) overstory and understory vegetation, five treatments were initiated in a 5–6-year-old longleaf pine stand: check, biennial arborescent plant control by directed herbicide application, and biennial burning in March, May, or July. The herbicide or prescribed fire treatments were applied in 1999, 2001, 2003, and 2005. All prescribed fires were intense and averaged 700 kJ/s/m of fire front across all 12 burns. Using pretreatment variables as covariates, longleaf pine survival and volume per hectare were significantly less on the three prescribed fire treatments than on checks. Least-square means in 2006 for survival were 70, 65, 64, 58, and 56% and volume per hectare was 129, 125, 65, 84, and 80 m³/ha on the check, herbicide, March-, May-, and July-burn treatments, respectively. A wildfire in March 2007 disproportionately killed pine trees on the study plots. In October 2007, pine volume per hectare was 85, 111, 68, 98, and 93 m³/ha and survival was 32, 41, 53, 57, and 55% on the check, herbicide, March-, May-, and July-burn treatments, respectively, after dropping trees that died through January 2009 from the database. Understory plant cover was also affected by treatment and the ensuing wildfire. In September 2006, herbaceous plant cover averaged 4% on the two unburned treatments and 42% on the three prescribed fire treatments. Seven months after the wildfire, herbaceous plant cover averaged 42% on the two previously unburned treatments and 50% on the three prescribed fire treatments. Before the wildfire, understory tree cover was significantly greater on checks (15%) than on the other four treatments (1.3%), but understory tree cover was similar across all five treatments 7 months after the wildfire averaging 1.1%. The greater apparent intensity of the wildfire on the previously unburned treatments most likely resulted from a greater accumulation of fuels on the check and herbicide plots that also collectively had a higher caloric content than fuels on the biennially prescribed burned plots. These results showed the destructive force of wildfire to overstory trees in unburned longleaf pine stands while also demonstrating the rejuvenating effects of wildfire within herbaceous plant communities. They caution for careful reintroduction of prescribed fire even if fire was excluded for less than a decade.     

Treatment of oak decline using pressurized injection capsules of antifungal materials

Quercus ilex and Quercus suber trees growing at several sites in Extremadura, Western Spain, that were showing symptoms of oak decline were injected with potassium phosphonate, quinosol or carbendazim using a low-pressure method of trunk injection composed of a pressurized capsule system. A team of four people injected between 120 and 189 trees per day, depending upon the density of the undergrowth vegetation. This labour cost represented, approximately, 15–20% of the total cost of the treatment. The potassium phosphonate-injected trees showed a significant improvement in vegetative growth within 2 years of the injection treatment, and they also showed some recovery from the decline symptoms during the third year. Only one injection treatment of an average of 3.5 capsules (corresponding to 24.5 g phosphonic acid) per tree of approximately 36 cm in diameter, was necessary to reduce the disease severity significantly. Indirectly, these results corroborate the implication of Phytophthora cinnamomi in oak decline within Spanish Quercus woodlands.     

Predicting understory maximum shrubs cover using altitude and overstory basal area in different Mediterranean forests

In some areas of the Mediterranean basin where the understory stratum represents a critical fire hazard, managing the canopy cover to control the understory shrubby vegetation is an ecological alternative to the current mechanical management techniques. In this study, we determine the relationship between the overstory basal area and the cover of the understory shrubby vegetation for different dominant canopy species (Pinaceae and Fagaceae species) along a wide altitudinal gradient in the province of Catalonia (Spain). Analyses were conducted using data from the Spanish National Forest Inventory. At the regional scale, when all stands are analysed together, a strong negative relationship between mean shrub cover and site elevation was found. Among the Pinaceae species, we found fairly good relationships between stand basal area and the maximum development of the shrub stratum for species located at intermediate elevations (Pinus nigra, Pinus sylvestris). However, at the extremes of the elevation-climatic gradient (Pinus halepensis and Pinus uncinata stands), stand basal area explained very little of the shrub cover variation probably because microsite and topographic factors override its effect. Among the Fagaceae species, a negative relationship between basal area and the maximum development of the shrub stratum was found in Quercus humilis and Fagus sylvatica dominated stands but not in Quercus ilex. This can be due to the particular canopy structure and management history of Q. ilex stands. In conclusion, our study revealed a marked effect of the tree layer composition and the environment on the relationship between the development of the understory and overstory tree structure. More fine-grained studies are needed to provide forest managers with more detailed information about the relationship between these two forest strata.     

Environmental and demographic correlates of tree recruitment and mortality in north Australian savannas

Tropical savannas cover approximately 20% of the earth’s land area, and therefore represent an important carbon store. Under scenarios of future climate change it is thus important to understand the demographic processes determining tree cover, namely tree recruitment, growth and mortality. This study measured tree recruitment and mortality in 123 (0.08 h) plots in Kakadu, Nitmiluk and Litchfield National Parks, in the Australian monsoonal tropics, over two consecutive 5-year intervals. Plots were located in two important habitats, both dominated by eucalyptus—lowland savanna and savanna growing on sandstone plateaux. All trees with diameter at breast height (DBH) ≥5 cm were tagged and identified. Recruitment was calculated as the proportion of tagged trees present at the end of an interval that were not present at the beginning. There were a total of 6666 and 6571 tree-intervals for mortality and recruitment, respectively. We used Akaike Information Criterion (AIC)-based model selection and multi-model inference to relate tree mortality and recruitment to fire frequency, mean annual rainfall (MAR), stand basal area, tree density and eco-taxonomic group. Recruitment decreased with tree density in both savanna types, and in lowland savanna, with the frequency of fires. In sandstone savanna, recruitment increased with MAR. Effects of fire on recruitment were better explained by season than severity of fire, while fire severity had a stronger influence on mortality. Mortality decreased with tree size up to about 25 cm DBH, but increased sharply when DBH exceeded 50 cm. Mortality increased with stand basal area, and increased with the frequency of late dry season fires in lowland savanna only. There was little evidence that mortality was affected by the frequency of early dry season fires or MAR. Both recruitment and mortality rates were higher for Acacia and Proteaceae species than for pantropical or Myrtaceae (including Eucalyptus) species. We identified several negative feedbacks, mediated by changes in tree density and stand basal area that help confer long-term stability to savanna tree cover. Nonetheless, changes such as a long-term increase in MAR or an increase in frequency or severity of fires are likely to result in changes in tree density, stand basal area and therefore carbon storage potential of savannas.     

Forest fire effects in beech dominated mountain forest of Iran

Most of world's forests of different climates have a history of fire, but with different severities. Fire regimes for broadleaf deciduous forests have return intervals that vary from many decades (or less) to centuries (or more). Iran has a total of 1.2 million ha of temperate forest in the north, where fires burn about 300–400 ha annually. This study focused on the impact of fire on forest structure, tree species quality, and regeneration composition (specially beech) in the Chelir forest of northern Iran. The results showed that forest fires changed the structure and had different effects on tree species composition between burned and control areas. Thin barked species such as oriental beech (Fagus orientalis Lipsky) and coliseum maple (Acer cappadocicum Gled.) have been affected more than those with thick bark, like hornbeam (Carpinus betulus L.) and chestnut-leaved oak (Quercus castaneifolia C.A. Mey). The density of oriental beech regeneration in the unburned area was greater than in the burned area, while the quantity of regeneration of hornbeam, coliseum maple and velvet maple (Acer velutinum Boiss) was higher in burned area. Forest fire had a greater effect on oriental beech quality, and changed regeneration composition in the burned area. Fire prevention activities should be considered as a silvicultural treatment for preserving these valuable forests.     

Fire resistance of tree species explains historical gallery forest community composition

Fire, climatic variability, and grazing by large herbivores have historically limited woody vegetation in the tallgrass prairie region of North America to gallery forests in protected areas along rivers and streams. Fire, in particular, has been a strong selective pressure against woody vegetation. Consequently, we expect that dominant tree species in these forests have developed mechanisms for tolerating periodic surface fires. Susceptibility of trees to fire damage depends in part on key properties of bark which influence heat transfer to the vascular cambium, including thickness, density, and moisture content. An historical (1983) survey of Konza Prairie Biological Station in northeast Kansas, USA indicated that gallery forests were co-dominated by Quercus macrocarpa and Quercus muehlenbergii, while Celtis occidentalis occurred as an important sub-dominant species. Populus deltoides, Gleditsia triacanthos, and Juniperus virginiana were relatively uncommon. To test the hypothesis that historically dominant gallery forest tree species are more resistant to fire damage than uncommon species, fire was applied to the bark of 10 individuals of each of these six species under conditions mimicking surface fires (400 °C for 120 s). Maximum temperature at the vascular cambium, bark thickness, bark moisture content, and bark density were measured. Trees were considered fire-resistant if the vascular cambium temperature remained below the thermal cell death threshold, 60 °C, throughout the treatment. Using logistic regression, bark thickness was found to be a significant predictor of lethal cambium temperatures (P = 0.002), while neither bark density nor moisture content were significantly related to lethal cambium temperature (P = 0.279 and P = 0.131, respectively). Across species, a minimum bark thickness of approximately 8.6 mm was necessary to maintain the vascular cambium temperature below 60 °C. Trees that produce thick bark quickly in juvenile size classes (P. deltoides, Q. macrocarpa, and Q. muehlenbergii) experienced lower temperatures at the vascular cambium than those which do not develop thick bark with increasing diameter (C. occidentalis, G. triacanthos, and J. virginiana). Ranking these tree species by either the DBH or age needed to develop the minimum protective bark thickness largely agreed with ranking based on historical relative importance. As fire frequency and intensity decrease in remnant tallgrass prairie of North America as a result of habitat fragmentation, fire suppression, and changing land management, fire-sensitive species may increase in relative importance in gallery forests because of increased juvenile survival.     

Positive fire–grass feedback in Mediterranean Basin woodlands

Fires can mediate switches between alternative vegetation types which may be more flammable and thus reinforce fire spread. We tested if there is a positive feedback between the expansion of the tussock grass Ampelodesmos mauritanica (hereafter Ampelodesmos) and fire hazard in Mediterranean Basin communities and its relation to tree cover decline. The effect of fire on Ampelodesmos population structure was analysed by surveying stands burned at different fire frequencies. The effect of vegetation dominated by Ampelodesmos on fire behaviour compared to other species coexisting in the community was predicted by the Rothermel fire propagation model BEHAVE. There was a negative correlation between pine cover and percentage of Ampelodesmos plants. Ampelodesmos mortality after fire is very low. Recently burned stands had a higher proportion of reproductive plants and higher seedling density than unburned stands. The high temperatures reached during fire may kill seeds, the higher seedling recruitment results from fast resprouting and higher seed production of burned plants compared to unburned plants 1 year after fire. Simulations with the BEHAVE fire model predict that Ampelodesmos increases fire intensity and spread because of its high accumulation of fuel load and standing dead material. The results suggest that there is a positive relationship between Ampelodesmos abundance and fire regime which increases the invasive potential of this grass and the fire risk of the community where it dominates.