Balancing Bioenergy and Biosecurity Policies: Estimating Current and Future Climate Suitability Patterns for a Bioenergy Crop

GCB Bioenergy (2014) 6, 587–598, doi: 10.1111/gcbb.12068

Balancing bioenergy and biosecurity policies: estimating current and future climate suitability patterns for a bioenergy crop

D. J. KRITICOS*,† ,H.T.MURPHY‡ , T. JOVANOVIC*, J. TAYLOR§ ,A.HERR*,J.RAISON* and D. O’CONNELL* *CSIRO Ecosystem Sciences, GPO Box 1700, Canberra, ACT, 2601, Australia, †EH Graham Centre for Agricultural Innovation, Charles Sturt University, Wagga Wagga, NSW, Australia, ‡CSIRO Ecosystem Sciences, PO Box 780, Atherton, QLD, Australia, §CSIRO Ecosystem Sciences, Box 312, Clayton South, VIC, 3169, Australia

Abstract In an apparent paradox, bioenergy crops offer potential benefits to a world adjusting to the challenges of climate change and declining fossil fuel stocks, as well as potential ecological and economic threats resulting from biological invasions. In considering this paradox it is important to understand that benefits and threats may not always be apparent in equal measure throughout the potential range of each candidate biofuel species. In some environments, a species could potentially produce valuable biological materials without posing a significant invasion threat. In this study, we develop a bioclimatic niche model for a candidate biofuel crop, Millettia pinnat- a, and apply the model to different climatic and irrigation scenarios to estimate the current and future patterns of climate suitability for its growth and naturalization. We use Australia as a case study for interpreting the niche model in terms that may be informative for both biofuels proponents and biosecurity regulators to plan management programmes that reflect the invasive potential in different areas. The model suggests that suitable growing conditions for M. pinnata in Australia are naturally restricted to the moist and semimoist tropics. Irriga- tion can extend the suitable growing conditions more widely throughout the tropics, and into more arid regions. Under future climate scenarios, suitable growing conditions for M. pinnata under natural rainfall contract towards the east coast, and extend southward into the subtropics. With irrigation, M. pinnata appears to have the potential in the future to naturalize across much of Australia. The bioclimatic modelling method demonstrated here is comparatively quick and easy, and can produce a rich array of data products to inform the inter- ests of both bioenergy proponents and biosecurity regulators. We show how this modelling can support the development of spatially explicit biosecurity policies designed to manage invasion risks in a manner that balances bioenergy and biosecurity concerns.

Keywords: biosecurity, CLIMEX, invasive alien species, Millettia pinnata, niche model, oilseed tree, Pongamia pinnata

Received 16 January 2013 and accepted 14 February 2013

distillation plants. Such efforts are intended to replace Introduction supplies of liquid fossil fuels and reduce the net emis- Presently, there is a great deal of interest worldwide in sions of GHGs. the development of biofuels industries as a replace- Much of the early interest in biofuels was focused ini- ment for fossil fuels. Demand for liquid fuels continues tially on using traditional food crops such as soybean, to rise at the same time as global production of liquid maize, palm oil and sugar cane to generate nonfood fossil fuels is set to decline (so-called peak oil), and products such as oil and ethanol. In the face of concern efforts to stem global emissions of greenhouse gases over food security issues posed by competition for land (GHGs) increase (Barney & Ditomaso, 2008; Penuelas~ & (Boddiger, 2007; Johansson & Azar, 2007; Dale et al., Carnicer, 2010; Verbruggen & Al Marchohi, 2010; 2010) and phosphorous fertilizers (Cordell et al., 2009), Farine et al., 2011). Biofuel proponents are interested in more recent fuel stock attention has been focused on establishing crop plantations to generate raw biological plant species that are not traditional crops. Traditional materials as inputs to new energy or liquid fuel crop systems are characterized generally by signiﬁcant bodies of knowledge regarding the plant’s growing Correspondence: D. J. Kriticos, tel. +61 2 6246 4252, requirements, production capacity under various fax +61 2 6246 4094, e-mail: [email protected] environmental conditions, weed potential, postharvest

© 2013 Blackwell Publishing Ltd 587 588 D. J. KRITICOS et al. processing systems and product qualities. For such The paradox here is that the knowledge of the spatial cropping systems it is relatively easy to develop suffi- patterns of production parameters and invasion risks ciently accurate estimates of the variables needed to needed to locate plantings and associated processing plan and implement a broad scale extension of the infrastructure accurately would be best gauged follow- geographical area under cultivation. Conversely, there ing the establishment of the industry, and from wide- is very little known about most of the noncrop plant spread trial sites. The problem of estimating a priori the species being proposed for biofuel feedstocks, except production potential, mirrors that of pest risk assess- that many of them are notorious weeds (Raghu et al., ment: the impacts of an unwanted invasive alien species 2006; Barney & Ditomaso, 2008; Di Tomaso et al., 2010). (IAS) can most reliably be gauged after it has invaded a The lack of knowledge surrounding the prospective new risk area, but this may then have little informative biofuels crops adds weight to concerns regarding both value for deciding how much effort should be expended their financial viability and the environmental impacts on preventing its further spread into that new risk area. of attempts to utilize marginal lands that are subse- Because biological invasions are generally irreversible, quently abandoned (Gutierrez & Ponti, 2009). Such con- modelling tools have been developed to estimate the cerns obviously parallel those regarding the financial potential for pests to spread and grow as a function of risks associated with the substantial investments climate. These tools can inform biosecurity authorities required to create the germplasm, planting, growing, of the risks posed by IAS either prior to their establish- harvesting, transportation and processing infrastructure ment in a jurisdiction, or soon thereafter (Eyre et al., required to support such embryonic industries. In estab- 2012). lishing a nascent biofuels industry in Europe based CLIMEX [Hearne Scientific Software, Melbourne, upon Miscanthus 9 giganteus, the main production limi- Australia, (Sutherst & Maywald, 1985; Sutherst et al., tations concerned poor overwintering and insufficient 2007)] is a popular computer package for pest risk water supply at some sites (Lewandowski et al., 2000). modelling of arthropods, diseases and weeds, with Prior knowledge of such climatic growth responses and more than 200 published applications. It has also been stresses in different areas could assist in planning field used successfully to project the potential distribution trials, and will allow the need for special agronomic of invasive plants including tropical woody weeds techniques and new genotypes with improved charac- (Kriticos et al., 2003a,b) and herbs (McConnachie et al., teristics to be factored into regional development plans 2011), subtropical trees (Watt et al., 2009), temperate (Barney & Di Tomaso, 2011). shrubs (Potter et al., 2009; Kriticos et al., 2011; Taylor The plant attributes that make for ideal biofuel crops et al., 2012) and grasses (Chejara et al., 2010; Watt et al., are frequently the same attributes that make for a suc- 2011; Bourdot^ et al., 2012). More recently, CLIMEX has cessful invasive plant (Raghu et al., 2006), including been used to assess the climatic niche for many climatic similarity to the introduced range (Barney & Di leading bioenergy crops (Barney & Di Tomaso, 2010, Tomaso, 2011). Such is the concern that biofuels crops 2011). could impact significantly on biodiversity and other The Compare Locations module in CLIMEX models conservation values that the International Union for climatic suitability for a species, rather than habitat Conservation of Nature (IUCN) is drafting guidelines similarity (Kriticos et al., 2007; Sutherst et al., 2007). for sustainable biofuels production that address the use CLIMEX uses biologically relevant functions to relate of invasive alien species; specifically, under Criterion species climate suitability to raw climate data. The spe- 7.e ‘Biofuel production shall not use crops considered as cies response functions in CLIMEX are founded in fun- alien invasive species under local conditions’ (Anony- damental ecological concepts. Because it characterizes mous, 2009b). Under this criterion, if a biofuels propo- the species response functions in a biologically mean- nent intends to use a species recorded in the Global ingful manner (Austin, 2007), it can project species Invasive Species Database (GISD), the operator needs to potential distributions into novel climates with more complete a risk assessment of the invasiveness of the confidence than empirical regression-based species dis- species in the local context. If there is clear evidence that tribution models (Kriticos & Randall, 2001; Webber the proposed species is likely to be invasive in the local et al., 2011), which are inherently risky when faced with context, the operator should not use the species. this extrapolation problem (Peterson et al., 2007; Duncan Although these draft guidelines do not yet carry the et al., 2009; Elith et al., 2010). In addition to using distri- weight of law, they indicate a set of principles and pro- bution records to estimate the potential geographical cedures that may be adopted widely by conservation range within which a species can persist, CLIMEX also and biosecurity authorities, and hence a hurdle that calculates weekly and annual growth indices based on industry proponents should prepare for during the the temperature and soil moisture regimes at each project planning and feasibility assessment stages. climate station.

Millettia pinnata (syn. Pongamia pinnata) is a medium- it should not be grown outside its natural range close to sized, leguminous tree belonging to the family Faba- national parks or watercourses. It should be declared a ceae. It is native to the Indian subcontinent and south- restricted plant that should not be grown near sensitive east Asia, and is often reported as naturalized or doubt- areas’. Clearly, based on the experience with J. curcas, fully native in Australia, China, Malaysia, Indonesia, proponents of M. pinnata as a biofuels feedstock in Aus- Japan, Vietnam and the United States. In India, M. pin- tralia will need to address concerns about its invasive nata has traditionally been used to fuel lanterns and potential, and to include plans to manage those risks. cooking stoves, and as medicines, fodder, beauty prod- Given that these risks are not equal throughout the ucts, shade, and environmental protection and is cur- potential planting range, it is prudent to use a species rently of major interest for biofuel production (Scott niche model to map these risks so that management et al., 2008; Kazakoff et al., 2010; Mukta & Sreevalli, plans can be tailored spatially, considering the potential 2010). It has been reported as being cold- and possibly for production and invasion risk management. By con- drought deciduous (Orwa et al., 2009; Murphy et al., sidering the invasion risk in a spatial manner, it may be 2012). Trees begin producing seeds at 4–5 years of age, possible to reduce excessive industry costs in managing with full production from about 10 years of age. invasion risks from the fuelstock where such effort is Accounts of seed yield from India are highly variable, not warranted. À À but suggest an average range of 8–90 kg tree 1 yr 1 We are unaware of any previous attempts to use a À although seed yields as high as 300 kg tree 1 have been species niche model to inform the planning and feasibil- reported from specially selected high-yielding trees (Di- ity assessment phase of a biofuels project [although Bar- vakara et al., 2010; Rao et al., 2011). Seed production has ney & Di Tomaso (2011) address many of the same also been shown to be highly variable within and across research elements in more general terms]. In this study, trial plantations in Australia; very little is currently we attempt to answer the following questions for our understood about the relationships between genotype, case study species (M. pinnata) under current and future environment and yield in M. pinnata (Murphy et al., climates, with particular emphasis on the Australia con- 2012). Depending on the extraction method, M. pinnata text: seeds have been found to contain approximately 40% 1 Where could the crop grow? extractable oil by weight (Mukta & Sreevali, 2009; Mur- 2 How well can it grow in different areas? phy et al., 2012). 3 Where is it likely to become invasive? In Hawaii, where M. pinnata has naturalized, it is considered ‘high risk’ for invasion (Buddenhagen et al., 2009; Pacific Island Ecosystems at Risk, 2009) based on a Materials and methods modified version of the Australian and New Zealand weed risk assessment system (Daehler et al., 2004). CLIMEX model However, the species is not yet listed on the Global CLIMEX is a dynamic model that integrates the weekly Invasive Species Database (http://www.issg.org/, last responses of a population to climate using a series of annual accessed 7 June 2012). indices. CLIMEX uses an annual growth index (GIA)to The Queensland State Government recently consid- describe the potential for population growth as a function of ered the risk and benefits of proposed widespread soil moisture and temperature during favourable conditions, planting of Jatropha curcas as a biofuels feedstock. and up to eight stress indices (cold, wet, hot, dry, cold-wet, Although J. cucas was perceived to be not as weedy as cold-dry, hot-wet and hot-dry) to determine the probability its congener J. gossypiifolia, the Queensland Government that the population can survive unfavourable conditions. The decided that sufficient risks existed to call for an indus- temperature (TI) and soil moisture (MI) growth indices are try code of practice aimed at ‘monitoring plantations for defined by three-segment linear equations that describe species responses to environmental factors in accordance with Shel- escapes and destroying any that occur’ (Anonymous, ford’s Law of Tolerance (Shelford, 1963). TI and MI are multi- 2009a:4). A Biosecurity Queensland weed risk assess- plied together to give the weekly Growth Index (GIW), which is ment determined that M. pinnata posed a low risk to integrated to an annual index, (GIA) scaled between 0 and 100, Queensland based primarily on the fact that there is indicating the potential for population growth, ignoring the currently no evidence that the species has significant effects of stresses. This relationship accords with the Sprengel– negative impacts as a weed elsewhere in the world, and Liebig Law of the Minimum (reviewed in van der Ploeg et al., that it is considered naturalized in Queensland (Csurhes 1999). Each of the stress indices is defined by a threshold value & Hankamer, 2010). However, in a review on potential of temperature or soil moisture, beyond which stress accumu- weed issues with biofuel crops in Australia, Low & lates at a rate defined by the corresponding stress accumulation Booth (2008:25–26) suggested that ‘Because this plant rate. CLIMEX also includes a mechanism for defining the mini- has a demonstrated capacity to spread from cultivation, mum amount of thermal accumulation (number of day

© 2013 Blackwell Publishing Ltd, GCB Bioenergy, 6, 587–598 590 D. J. KRITICOS et al. degrees) during the growing season that is necessary for popu- georeferenced locations outside Australasia. In Australia, M. lation persistence (PDD). pinnata is found in the northern tropics and subtropical east The growth and stress indices are calculated weekly and are coast ranging from the coastal fringe around Darwin through then combined into an overall annual index of climatic suitabil- Cape York and as far south as northern New South Wales ity, the Ecoclimatic Index (EI), which gives a measure of the (Fig. 1). Records further inland are from locations along major potential of a given location to support a permanent population rivers including the Wenlock, Archer and Holroyd in the Cape of the species. The Ecoclimatic Index (EI), ranges from 0 for York area, and the Norman River in the Gulf of Carpentaria. It locations at which the species is not able to persist to 100 for is grown commonly as a street tree in and near Brisbane, locations that are optimal for the species year round. including the Sunshine Coast, although it has not been CLIMEX is a bioclimatic model, relying on a database of cli- reported as naturalizing there. A number of trial plantations matic variables of long-term monthly precipitation totals, aver- have been established throughout Queensland, in the Northern ages of minimum and maximum temperatures and averages of Territory and in Western Australia (Murphy et al., 2012). relative humidity at 09:00 and 15:00 hours. The historical Location records for M. pinnata were cleaned, removing climate data set used for these analyses was the CliMond data records that were located in geographically implausible loca- set (www.climond.org), with a spatial resolution of 10′, using tions, such as the sea. Geographical outliers were checked care- station records centred on 1975 (Kriticos et al., 2012). fully to verify that the geographical coordinates matched other location information in the records. Where there was a disagreement between the coordinates and the location descrip- Known distribution of Millettia pinnata and model tion, if the coordinates could be reasonably and confidently fitting corrected they were adjusted, otherwise the records were Herbarium records of the worldwide geographical distribution deleted. The cleaned records were transformed into shapefiles and imported into CLIMEX to aid in fitting the stress of Millettia pinnata and its synonym Pongamia pinnata were sourced from the Global Biodiversity Information Facility responses. The model was fitted to location records outside (GBIF) and Australia’s Virtual Herbarium web database. Loca- Australia, while the Australian data were reserved for model validation. Using the Compare Locations model in CLIMEX, tion records with anomalous geocoding whose geographical coordinates could not be established reliably were removed. the parameters for the Wet Tropics template, which most clo- M. pinnata Additional location records were sourced from journal publica- sely approximated the climatic of was adjusted until tions wherever geographical coordinates were given or could the distribution points fell within the area modelled as climatically suitable. Parameter values were also informed by several be established using sufficiently detailed locality information (Ramesh, 2008; Hooda et al., 2009; Divakara et al., 2010; Pandey knowledge domains, including field observations of survival et al., 2010; Rao et al., 2011). and ecophysiology. The native range of M. pinnata is variously described as India and Burma, or south-east Asia (Troup, 1983; Daniel, 1997; Stress indices Csurhes & Hankamer, 2010). The naturalized range of M. pinnata extends through Central America, southern USA, Africa, Cold stress. Although the stems of M. pinnata die back in the Asia and Australia (Fig. 1); although we could find no reported presence of frost, it has been observed to recover readily from

Fig. 1 Map of locations where Millettia pinnata has been recorded growing. Source: Global Biodiversity Information Facility (GBIF), Australia’s Virtual Herbarium records and locations extracted from (Ramesh, 2008; Hooda et al., 2009; Divakara et al., 2010; Pandey et al., 2010; Rao et al., 2011). These records may include locations where M. pinnata has not been noted as having naturalized.

© 2013 Blackwell Publishing Ltd, GCB Bioenergy, 6, 587–598 BALANCING BIOENERGY AND BIOSECURITY POLICIES 591 the effects of a one-off event. This agrees with observations that effects of otherwise lethal heat stress, and the promotion of it is cold-deciduous and tolerant of slight frosts (Prasad & Pan- rapid repair of cellular function during a subsequent (mild- dey, 1987; Mukta & Sreevalli, 2010; Murphy et al., 2012). A temperature) recovery period (Horvath et al., 2012). The critical growing degree-day cold stress function was fitted to the distri- elements for our modelling are that short-term survival of high bution data for M. pinnata. This two-parameter function temperatures may be a poor indicator of survival under sus- accounts for the necessity for the plant to grow a minimal tained exposure to high temperatures. Experiments with amount to balance respiration requirements. Unless the plant increasing TTHS resulted in a poorer model fit to the known experiences a minimum of 15 °C days per week (DTCS) above distribution. the lower temperature threshold for development (DV0, 15 °C), it is assumed to draw down on photosynthetic energy reserves. À The cold stress accumulation rate (DHCS) of À0.001 week 1 Growth indices allowed the plant to persist in Florida, the lower slopes of the Information on the ecophysiology of M. pinnata was used to Himalayas and the locations in south-western Taiwan. Given inform the choice of parameters for the growth index in CLI- the nature and extent of the degree-day cold stress, we consid- MEX. ered it unnecessary to include the effects of lethal cold stress associated with frosts. Temperature index. The lower temperature threshold for development was set to 15 °C. The lower and upper limits for Dry stress. Dry stress was assumed to commence as soil mois- growth (DV1 and DV2) were set to 28 °C and 33 °C, respec- ture drops below permanent wilting point, and so the dry tively, based on the response of similar tropical tree species. stress threshold (SMDS) was set to 0.1. The fitted stress accu- The maximum temperature for growth (DV3) was set to 36 °C. À À1 mulation rate of 0.04 week indicates that drought decidu- Taking into account the averaging effect, this accords well with ousness allows mature M. pinnata plants to survive moderately the maximum temperature in the area where M. pinnata occurs prolonged periods with inadequate soil moisture (less than in India, 38 °C (Mukta & Sreevalli, 2010). 6 months). These parameters correspond approximately to the reported rainfall limits for the species (600 mm rainfall and Moisture index. The lower limit for growth (SM0) was set to 2200 mm potential evaporation). Using these parameters, the 0.1, which for many soils and species approximates permanent Ecoclimatic Index limit for M. pinnata under natural rainfall wilting point. The optimum soil moisture levels for growth coincides with the most xeric known locations in India where it (SM1 and SM2) were set to biologically reasonable values has been observed, where trees are not restricted to roadsides (Table 1). The upper soil moisture level for growth (SM3) was and irrigation areas. set to 2.5, indicating that M. pinnata is moderately tolerant of Using these parameters, M. pinnata is modelled as unable to water-logging. persist at xeric locations such as Sirsa (29°31′N,75°01′E) in India, or Kunanurra (15°46′S,128°44′E) in Western Australia without the aid of irrigation during the prolonged dry season. Growing degree-days Using Google Earth, the outlying Indian records are located in areas where agricultural irrigation is apparently practised, and Millettia pinnata is not found above 600 m in the foothills of the trees are restricted to roadsides and other favourable habitats. Himalaya Mountains (Troup, 1983). This region represents a clear naturalization limit that is likely to be due to climatic factors. Because the plant is cultivated in cooler conditions than Wet stress. Noting that M. pinnata is moderately tolerant of are experienced at 600 m in this region, it is likely that this lim- water-logging (Troup, 1983), the threshold for wet stress itation is related to the annual heat sum required to produce (SMWS) was set to 2.5, and the accumulation rate (HWS) was À viable seeds. To estimate the growing degree sum that is asso- set to 0.002 week 1. ciated with this limit, CLIMEX was run with the DV0 parameter set at 15 °C using the CliMond climate data set (Kriticos Heat stress. The heat stress threshold TTHS was set to 37 °C, et al., 2012). The data were exported to ArcGIS V9.3 (ESRI, Red- marginally above the maximum temperature for growth (DV3), lands, CA, USA), and points were extracted from the base of and the stress accumulation rate (THHS) was fitted to allow the Himalayas where the altitude was between 500 and 700 m. survival at the hottest known suitable location in India. Orwa The corresponding average growing degree-days above 15 °C et al. (2009) claim that mature trees of M. pinnata can withstand were then plotted against altitude (Fig. 2). The threshold heat temperatures of 50 °C, and Murphy et al. (2012) report that sum immediately below 600 m is approximately 2850 °C days. plants can survive a short duration exposure to 65 °C when This value was used in the CLIMEX model to indicate the min- well-irrigated. However, it is clear that ‘For all living organ- imum annual heat sum required by M. pinnata for naturaliza- isms, temperatures only moderately above the respective opti- tion. mum growth temperature represent a challenging problem for Setting the PDD value to 2850 resulted in four location survival’ (Richter et al., 2010:253). Heat shock response is a records in Papua New Guinea appearing to fall in areas too rapid and transient energy-demanding gene-expression system. cold for persistence. These records were found to have precise Exposure to a mild sublethal temperature stress may lead to location coordinates that placed them at the bottom of deep acquired thermotolerance through the induction of heat shock ravines. The climate in the corresponding climate cells were proteins; conferring short-term protection from the damaging invariably at higher elevation than the location records.

Table 1 CLIMEX parameter values used for Millettia pinnata. Parameter mnemonics taken from Sutherst et al. (2007)

Index Parameter Valuea

Temperature DV0 = lower threshold 15 °C DV1 = lower optimum temperature 28 °C DV2 = upper optimum temperature 33 °C DV3 = upper threshold 36 °C Moisture SM0 = lower soil moisture threshold 0.1 SM1 = lower optimum soil moisture 0.7 SM2 = upper optimum soil moisture 1.5 SM3 = upper soil moisture threshold 2.5 Cold stress DTCS = growing degree-day threshold 15 °C Days À THCS = stress accumulation rate À0.001 Week 1 Heat stress TTHS = temperature threshold 37 °C À THHS = stress accumulation rate 0.0002 Week 1 Dry stress SMDS = threshold soil moisture 0.1 À HDS = stress accumulation rate À0.04 Week 1 Wet stress SMWS = threshold soil moisture 2.5 À HWS = stress accumulation rate 0.002 Week 1 Annual heat sum PDD = degree-day threshold 2900 °C Days aValues without units are dimensionless indices of a 100 mm single-bucket soil moisture model.

3300

3200

3100

3000

2900

2800

2700 Growing degree days

2600

2500

2400 450 500 550 600 650 700 750 Elevation (m)

Fig. 2 Heat sum above 15 °C at the foothills of the Himalayas. Growing degree-days were calculated using CLIMEX running the Cli- Mond 1975 historical data set V1 (Kriticos et al., 2012).

Another set of records in Taiwan were also found outside the constraints such as the population degree-day limit (PDD). The PDD limit. These records were all located in urban settings, potential for establishing plantations in areas outside the zone and so were presumed to be unnaturalized amenity plantings. where M. pinnata could establish were explored under a natu- We tested this hypothesis by relaxing the threshold to accom- ral rainfall scenario, and one including an irrigation scenario in À modate the outlying records and concluded that on the balance which 2.5 mm day 1 was applied as a top-up to natural rain- of evidence, the Himalayan limit was the most plausible. fall. That is, irrigation was only applied when natural rainfall did not meet this threshold, and then it was only used to bring Natural rainfall and irrigation scenarios the apparent rainfall up to this amount. Irrigated areas in Aus- tralia were extracted from national-scale land use mapping CLIMEX calculates a range of state variables, including the (ABARES, 2010). We buffered irrigation areas by 50 km to annual Growth Index (GIA). The GIA indicates the climatic establish broad zones where irrigation is practised, and there- potential for growth in the absence of stresses and other range fore is likely to be feasible, and to aid in visibility of these

© 2013 Blackwell Publishing Ltd, GCB Bioenergy, 6, 587–598 BALANCING BIOENERGY AND BIOSECURITY POLICIES 593 zones on the maps. It should be noted, however, that irrigation this information in IPCC (2000), and because we do not wish to infrastructure and access to irrigation water may not be wide- encourage transsubstantiation with respect to the scenario spread throughout these areas. results. Our application of the future climate scenario to the CLIMEX model is not intended to be in any sense predictive. Our intention is to use it to understand something of the sensi- Climate change tivity of our study system and any policy indicators to future climate changes, rather than as a data product in its own right The impacts of climate change on the potential for M. pinnata (O’Neill et al., 2008). Using the GCM data simply provides a to grow or pose an invasion risk were explored using a climate future climate scenario that is internally consistent. scenario model for 2080 taken from the CliMond data set (Kriti- cos et al., 2012). The selected climate data sets were developed using the high A2 emission scenario (IPCC, 2000) applied to Results the CSIRO Mk 3.0 global climate model. The A2 emissions scenario was selected because it remains a plausible future The global climate suitability for M. pinnata is indicated emission scenario, taking into account recent observations of in Fig. 3. The potential naturalized range of M. pinnata greenhouse gas emissions and rising temperatures (Rahmstorf [where it can produce seed viable for natural germina- et al., 2007; Raupach et al., 2007). We have not included details tion, indicated by the Ecoclimatic Index (EI) greater than of the A2 scenario here because interested readers can access 0] is restricted to the moist and semimoist tropics.

(a)

(b)

Fig. 3 Global climate suitability (EI) for Milletia pinnata under (a) current climatic conditions and (b) modelled climatic conditions for 2080 based on the high A2 emission scenario applied to the CSIRO Mk 3.0 global climate model. Distribution records for M. pinnata (as per Fig. 1) are represented by blue circles.

Besides its native range in India and Bangladesh, its east coast of Australia and extends further southwards potential naturalized range extends into Florida and into south-east Queensland and northern New South marginally into Texas in North America, the Caribbean Wales (Fig. 4e). Only a small area in north-east North- and the warmer low-lying parts of Central America, ern Territory remains suitable for persistence under nat- most of Brazil and the warmer parts of South America, ural rainfall scenarios. With irrigation, the climatically the lower-lying parts of Sub-Saharan Africa as far south suitable area extends further to the south (Fig. 4f). In as northern Angola in the west and Mozambique in the 2080 under irrigation, the existing plantations at Ku- east. Most of south-east Asia is highly suitable except nunurra and near Roma in south-west Queensland are for the highlands of Irian Jaya and Papua New Guinea reduced to marginal climatic suitability, as is much of (too cold), and most of Australia (too dry or too cold). the interior of Australia, due primarily to heat stress. In Only the southernmost parts of China and some south- 2080, the existing plantations in south-east Queensland western parts of Taiwan appear presently suitable for shift from a climate where naturalization is unlikely M. pinnata establishment. (EI < 1) to one where the naturalization potential is In Australia, the EI indicates that M. pinnata has the increased (Fig. 5). potential to persist naturally throughout the coastal fringe of tropical Australia where adequate rainfall and Discussion a climate that is relatively free of frosts provide the most suitable growing regions (Fig. 4a, Fig. 5). Potential range and invasion risks The nonplantation distribution records in Australia south of the area indicated as climatically suitable for The CLIMEX niche model indicates that M. pinnata has naturalization, consist of amenity plantings that are pres- the potential to establish under natural rainfall condi- ently unlikely to be capable of naturalizing. For example, tions throughout the moist tropics (Fig. 3a), highlighting a small group of trees were reported growing in the Aus- invasion and production potential in large regions of tralian Botanic Garden, Mount Annan in Sydney, New Central and South America and Africa. Interestingly, South Wales Australia (Armstrong, personal communi- the climatic conditions in these regions appear even cation, 2010). The staff at the garden confirmed that the more suitable for M. pinnata than in the native range plants had produced seeds, but had not naturalized. (Fig. 3a), due to the reduced seasonality, and hence Similarly, street trees in Brisbane and at the Sunshine reduced stresses in equatorial regions. In the future cli- Coast had not been reported as naturalizing, including mate scenario, there appears to be only modest potential some plants in a cemetery with little adjacent distur- for range expansion (Fig. 3b). bance (P. Gresshoff personal communication, 2010). The In Australia, M. pinnata’s potential range under his- model is therefore logically consistent with presence of torical climatic conditions extends over a relatively lim- these location records in the zone with a positive GIA ited area of the north-east. The modelled range is and an EI value of zero. Many of the present plantations consistent with its documented naturalized range here, in Australia are located in the zone modelled as having a although the modelled potential range extends further positive GIA, (Fig. 4c) but an inadequate PDD. inland into central Queensland (Fig. 4a), perhaps indi- Application of the irrigation scenario significantly cating an invasion lag or a genetic bottleneck distin- increases the area where M. pinnata has the potential to guishing the Australian genotype from the Indian persist across northern Australia (Fig. 4b). For example, populations that occupy climates similar to central the model indicates that irrigation may be necessary to north Queensland. Favourable climatic environments in > allow plants to survive the prolonged dry season at the terms of population growth (GIA 0) are, however, plantation site in Kununurra, Western Australia more extensive, and occur throughout Australia except (Fig. 4b). However, it may not be practical or feasible to in the arid interior (Fig. 4b). In these areas, M. pinnata irrigate over much of this extent; currently irrigated may be able to survive and grow under natural climatic land use in northern Australia is limited to very local- conditions but it is unlikely to reproduce in most years ized areas around Kununurra in Western Australia, due to insufficient heat accumulation for physiological Darwin and Katherine in the Northern Territory and development or insufficient soil moisture. Thus, the small patches in north-west Queensland (hatched areas potential for M. pinnata to invade and establish outside on Fig. 4 and 5). plantations in these areas is likely limited. Biofuels industry proponents with M. pinnata plantations in south-east Queensland have noted that irriga- Climate change tion is generally required or advisable during the first Under the 2080 climate scenario, the area in which EI is 7 years following establishment to ensure survival of greater than zero contracts to a narrow band along the seedlings (Murphy et al., 2012); none of these plantations

(a) (b)

(e) (f)

Fig. 4 Climatic suitability (EI) and Growth Potential (GIA) in Australia for Millettia pinnata under current and future climate scenar- – – ios. (a) EI natural rainfall conditions current climate, (b) EI irrigated conditions current climate, (c) GIA natural rainfall conditions – – – current climate, (d) GIA irrigated conditions current climate, (e) EI natural rainfall conditions 2080 A2 scenario, and (f) EI irrigated conditions – 2080 A2 scenario. Blue dots indicate ﬁeld distribution records for M. pinnata, blue stars indicate M. pinnata plantations, irrigated where necessary.

(a) Australia, particularly in the zones indicated as having naturalization potential under the irrigation scenario (Fig. 5), may provide suitable conditions for both establishment and persistence. The level of climate similarity with the native range of a species is a consistent indicator of likely invasion success in a new environment, along with weediness elsewhere and propagule pressure (Hayes & Barry, 2008). There are some regions in northern Australia where climatic similarity with the native range is high under either natural or irrigated conditions. While not currently listed on the Global Invasive Species Database, M. pinnata is considered a high-risk weed in Hawaii, and has demonstrated the capacity to naturalize there (Buddenhagen et al., 2009). In theory, propagule pressure from M. pinnata plantations may be low as seeds are harvested mechanically from the trees for oil extrac- (b) tion. However, widespread planting will increase the risk of escape from cultivation, and the introduction of germplasm from more vigorous, high-yielding individu- als and elite stock from India or elsewhere into large- scale plantations may increase the risk of invasiveness significantly (Low & Booth, 2008; Murphy et al., 2012). Furthermore, it is almost inevitable that some plantations will be abandoned over time for any number of reasons. Given the expected lifespan of M. pinnata (50– 100 years) it seems prudent to assess the invasion risk over (at least) the timeframes modelled here (70 years), taking into account sensitivity to potential climate changes and other regional and continental changes (e.g., expansion of irrigation infrastructure in northern Australia). Plantations established today in areas where Fig. 5 Modelled naturalization potential across Australia for naturalization potential is unlikely, for example, in Millettia pinnata under natural and irrigated conditions under south-east Queensland, may experience increasingly (a) current climate, and (b) future climate (2080 A2) scenario. Naturalization can occur where EI > 0 under natural or irri- favourable conditions over their lifespan, and trees may gated conditions; naturalization is unlikely where EI = 0. There reach reproductive peak in conditions conducive to nat- are no areas in Australia where GI < 1 under irrigated condi- uralization. Therefore, it may be wise for any plantations. Dots indicate field distribution records for M. pinnata, tions established in areas noted as ‘naturalisation stars indicate M. pinnata plantations. potential’ and ‘naturalisation potential (irrigation)’ (Fig. 5) under current or future climate scenarios to have reached the production phase, so it is unknown implement monitoring and mitigation procedures to whether irrigation will be required on an ongoing basis reduce the risk of M. pinnata establishing beyond the to ensure commercially attractive yields. Under irriga- plantation. tion, the potential for self-propagation is significantly Clearly, prudent risk management requires a balanc- more widespread, extending across northern Australia. ing of risk reduction and benefit maximization consider- Practical considerations associated with infrastructure ations. In the case of M. pinnata this means exploring and water availability will most likely greatly reduce the potential for economically attractive plantations, the potential for plantation establishment over much of while managing the invasion potential in those areas this area (Fig. 5). Of particular importance in terms of where it might naturalize. The role of cultivation in the invasion potential is that M. pinnata may be able to nat- spread of invasive alien plants (Mack & Lonsdale, 2001; uralize and persist in microsites with favourable soil Emms et al., 2005) is sufficiently well understood that moisture that are otherwise projected to be unsuitable many regulators are wary of plans to establish planta- or marginal in the absence of irrigation. Gullies, creeks, tions of alien species. Because M. pinnata can be grown riparian zones and seepage lines across northern in areas beyond its climatic naturalization potential,

© 2013 Blackwell Publishing Ltd, GCB Bioenergy, 6, 587–598 BALANCING BIOENERGY AND BIOSECURITY POLICIES 597 there is the opportunity to use knowledge of its ecology estimates of regional oil yield potentially suitable for to segregate policies and regulations spatially, in such a infrastructure planning purposes. way that the public and private benefits are balanced. The model results highlight those areas where, under Conclusion historical and future climate scenarios there is a potential for M. pinnata to establish and become invasive, and The modelling and analytical method demonstrated those areas where it may be cultivated. Biosecurity here was relatively quick to generate. It allowed us to agencies and industry proponents can use this informa- synthesize information from a diverse range of knowl- tion to decide whether any plantations within that zone edge domains (biology, ecology, geography and climate deserve measures to mitigate the invasion threats. change) to estimate this species’ potential for establish- The climate change scenario used in this study is just ment and growth. The method can produce a rich array that, a scenario. It does not represent data about the of data products to inform the risk assessments required future, but rather, provides us with an internally consis- under the draft IUCN guidelines on the use of invasive tent means of exploring the sensitivity of our biophysi- species for biofuels (Anonymous, 2009b), and possibly cal system to global climate change. There are deep and also help inform potential industry operators of the irreducible uncertainties regarding our future climate, potential for plantation production and climatic chal- stemming primarily from our inability to forecast future lenges in different regions. This method may also be socio-economic development patterns. These uncertain- used to develop spatially explicit biosecurity policies ties mean that scenarios such as the one used in Fig. 5b designed to manage invasion risks in a more sophisti- should not be used as the strict basis for planning and cated, socially responsible manner that allows energy policy development, but rather, should be used to indi- production and manages biosecurity concerns. cate the direction in which the biophysical system could plausibly tend in the future. This modelling can there- References fore be used to indicate broadly, areas where risks may change in the future, and the manner in which they are ABARES (2010) Land Use of Australia, Version 4, 2005/2006 (September 2010 release). Department of Agriculture Fisheries and Forestry (Australian Bureau of Agricul- likely to change. In accord with the original purpose for tural and Resource Economics and Sciences). http://adl.brs.gov.au/anrdl/meta- developing scenario analyses, we can use this informa- data_files/pa_luav4g9abl07811a00.xml. tion to assess the robustness of our management poli- Anonymous (2009a) Jatropha curcas Production in Queensland: A Position Paper. Department of Employment Economic Development and Innovation, Brisbane. cies and systems (Van der Heijden, 1997). Anonymous (2009b) Report From 2nd IUCN Biofuels and Invasives Workshop - Sum- Given the niche modelling and forecasting uncertain- mary. Managing Biodiversity Risks of Biofuels, Nairobi, Kenya, 5-6th October, ties regarding greenhouse gas emission patterns, in any 2009 Austin MP (2007) Species distribution models and ecological theory: a critical assess- biosecurity policy formulation, it may be wise to include ment and some possible new approaches. Ecological Modelling, 200,1–19. safety margins, and an ongoing surveillance protocol to Barney JN, Di Tomaso JM (2010) Bioclimatic predictions of habitat suitability for the check whether the CLIMEX model’s assumptions are biofuel switchgrass in North America under current and future climate scenarios. Biomass and Bioenergy, 34, 124–133. flawed regarding these limits, and to monitor how the Barney JN, Di Tomaso JM (2011) Global climate niche estimates for bioenergy crops risks are changing in the future in relation to realized and invasive species of agronomic origin: potential problems and opportunities. climatic changes. PLoS ONE, 6, e17222. Barney JN, Ditomaso JM (2008) Nonnative species and bioenergy: are we cultivating the next invader? BioScience, 58,64–70. Boddiger D (2007) Boosting biofuel crops could threaten food security. The Lancet, Production potential 370, 923–924. Bourdot^ GW, Lamoureaux SL, Watt MS, Kriticos DJ (2012) The potential global dis- Very little is currently understood about the relationship tribution of the invasive weed Nassella neesiana under current and future climates. between climate and seed production or environmental Biological Invasions, 14, 1545–1556. stress (induced by soil and climatic factors) and oil yield Buddenhagen CE, Chimera C, Clifford P (2009) Assessing biofuel crop invasiveness: a case study. PLoS ONE, 4, e5261. in M. pinnata. Seed production across, and within, trial Chejara VK, Kriticos DJ, Kristiansen P, Sindel BM, Whalley RDB, Nadolny C (2010) sites and on street trees in Australia is highly variable The current and future potential geographical distribution of Hyparrhenia hirta. – (Murphy et al., 2012). Industry proponents could use the Weed Research, 50, 174 184. Cordell D, Drangert J-O, White S (2009) The story of phosphorus: global food secu- model results as an indication of growth potential and rity and food for thought. Global Environmental Change, 19, 292–305. the need for, and value of irrigation in different areas. Csurhes S, Hankamer C (2010) Weed Risk Assessment: Pongamia. The State of Queens- This information could be used to guide industrial land. Department of Empolyment, Economic Development and Innovation, Bris- bane, Queensland, Australia. planning and the location of experimental plantings. Daehler CC, Denslow JS, Ansari S, Kuo HC (2004) A risk-assessment system for The model lends itself to the calibration of a meso-scale screening out invasive pest plants from Hawaii and other Pacific Islands. Conser- – function relating climate suitability to yield potential. vation Biology, 18, 360 368. Dale VH, Kline KL, Wiens J, Fargione J (2010) Biofuels: Implications for Land Use and When combined with a land availability and market Biodiversity. Ecological Society of America. Washington, DC. analysis, this same function could underpin robust

Daniel JN (1997) Pongamia pinnata – a nitrogen fixing tree for oilseed, NFT highlights. Murphy H, O’Connell D, Seaton G et al. (2012) A common view of the opportunities, NFTA, 97–03. challenges, and research actions for Pongamia in Australia. BioEnergy Research, 5, Di Tomaso J, Reaser J, Dionigi C, Doering O, Chilton E, Schardt J (2010) Biofuel vs 778–800. bioinvasion: seeding policy priorities. Environmental science & technology, 44, 6906– O’Neill B, Pulver S, VanDeveer S, Garb Y (2008) Where next with global environ- 6910. mental scenarios? Environmental Research Letters, 3, 045012. Divakara BN, Alur AS, Tripati S (2010) Genetic variability and relationship of pod Orwa C, Mutua A, Kindt R, Jamnadass R, Simons A (2009) Agroforestree Database: A and seed traits in Pongamia Pinnata (L.) Pierre., a potential agroforestry tree. Inter- Tree Reference and Selection Guide, version 4.0, World Agroforestry Centre, Kenya. national Journal of Plant Production, 4, 129–141. Pacific Island Ecosystems at Risk (2009) Weed Risk Assessment – Pongamia [Millettia] Duncan RP, Cassey P, Blackburn TM (2009) Do climate envelope models transfer? A pinnata. Department of Forestry, USDA. manipulative test using dung beetle introductions. Proceedings of the Royal Society Pandey AK, Nivedika G, Pankaj B, Sharma DS (2010) Evaluation of Pongamia pinnata B-Biological Sciences, 276, 1449–1457. (L.) Pierre. progenies for their growth performance in Madhya Pradesh, India. Elith J, Kearney M, Phillips S (2010) The art of modelling range-shifting species. World Applied Sciences Journal, 10, 225–233. Methods in Ecology and Evolution, 1, 330–342. Penuelas~ J, Carnicer J (2010) Climate change and peak oil: the urgent need for a tran- Emms J, Virtue JG, Preston C, Bellotti WD (2005) Legumes in temperate Australia: a sition to a non-carbon-emitting society. Ambio, 39,85–90. survey of naturalisation and impact in natural ecosystems. Biological Conservation, Peterson AT, Papes M, Eaton M (2007) Transferability and model evaluation in 125, 323–333. ecological niche modeling: a comparison of GARP and Maxent. Ecography, 30, Eyre D, Baker RHA, Brunel S et al. (2012) Rating and mapping the suitability of the 550–560. climate for pest risk analysis. EPPO Bulletin, 42,48–55. van der Ploeg RR, Bohm€ W, Kirkham MB (1999) On the origin of the theory of min- Farine DR, O’Connell DA, Raison RJ et al. (2011) An assessment of biomass for bio- eral nutrition of plants and the law of the minimum. Soil Science Society of America electricity and biofuel, and for greenhouse gas emission reduction in Australia. Journal, 63, 1055–1062. Global Change Biology Bioenergy, 4, 148–175. Potter KJB, Kriticos DJ, Watt MS, Leriche A (2009) The potential distribution of Gutierrez AP, Ponti L (2009) Bioeconomic sustainability of cellulosic biofuel produc- Scotch broom, Cytisus scoparius (L) Link: a weed of commercial forestry, pastoral tion on marginal lands. Bulletin of Science, Technology & Society, 29, 213–225. systems and natural ecosystems. Weed Research, 49, 271–282. Hayes KR, Barry SC (2008) Are there any consistent predictors of invasion success? Prasad R, Pandey RK (1987) Vegetation damage by frost in natural forests of Madha Biological Invasions, 10, 483–506. Pradesh. Journal of Tropical Forestry, 3, 273–278. Hooda MS, Dhillon RS, Dhanda S, Kumari S, Dalal V, Jattan M (2009) Genetic diver- Raghu S, Anderson RC, Daehler CC, Davis AS, Wiedenmann RN, Simberloff D, gence studies in plus trees of Pongamia pinnata (Karanj). Indian Forester, 135, 1069– Mack RN (2006) Adding biofuels to the invasive species fire? Science, 313, 1080. 1742. Horvath I, Glatz A, Nakamoto H et al. (2012) Heat shock response in photosynthetic Rahmstorf S, Cazenave A, Church JA, Hansen JE, Keeling RF, Parker DE, Somerville organisms: membrane and lipid connections. Progress in Lipid Research, 51, 208–220. RCJ (2007) Recent climate observations compared to projections. Science, 316, 709. IPCC (2000) Special Report on Emissions Scenarios: A Special Report of Working Group Iii Ramesh N (2008) Studies on Provenance, Nursery Mixture And Pre Sowing Treatments of The Intergovernmental Panel on Climate Change. Cambridge University Press, on Quality and Characterisation of Pongamia. MSc Thesis, Dharwad University of Cambridge, United Kingdom. Agricultural Sciences, Dharwad. Johansson D, Azar C (2007) A scenario based analysis of land competition between Rao G, Shanker A, Srinivas I, Korwar G, Venkateswarlu B (2011) Diversity and vari- food and bioenergy production in the US. Climatic Change, 82, 267–291. ability in seed characters and growth of Pongamia pinnata; (L.) Pierre accessions. Kazakoff SH, Gresshoff PM, Scott PT (2010) Pongamia pinnata, a sustainable feed- Trees - Structure and Function, 25, 725–734. stock for biodiesel production. In Energy Crops: RSC Energy and Environment Series Raupach MR, Marland G, Ciais P, Le Quere C, Canadell JG, Klepper G, Field CB No. 3 (eds Halford NG, Karp A), pp. 233–258. Royal Society of Chemistry. www. (2007) Global and regional drivers of accelerating CO2 emissions. Proceedings of rsc.org the National Academy of Sciences of the USA, 104, 10288–10293. Kriticos DJ, Randall RP (2001) A comparison of systems to analyse potential weed Richter K, Haslbeck M, Buchner J (2010) The heat shock response: life on the verge distributions. In: Weed Risk Assessment (eds Groves RH, Panetta FD, Virtue JG), of death. Molecular Cell, 40, 253–266. pp. 61–79. CSIRO Publishing, Melbourne, Australia. Scott PT, Pregelj L, Chen N, Hadler JS, Djordjevic MA, Gresshoff PM (2008) Pong- Kriticos DJ, Sutherst RW, Brown JR, Adkins SA, Maywald GF (2003a) Climate amia pinnata: an untapped resource for the biofuels industry of the future. BioEn- change and biotic invasions: a case history of a tropical woody vine. Biological ergy Research, 1,2–11. Invasions, 5, 145–165. Shelford VE (1963) The Ecology of North America. University of Illinois Press, Urbana, Kriticos DJ, Sutherst RW, Brown JR, Adkins SA, Maywald GF (2003b) Climate IL, USA. change and the potential distribution of an invasive alien plant: Acacia nilotica Sutherst RW, Maywald GF (1985) A computerised system for matching climates in ssp. indica in Australia. Journal of Applied Ecology, 40, 111–124. ecology. Agriculture, Ecosystems and Environment, 13, 281–299. Kriticos DJ, Stephens AEA, Leriche A (2007) Effect of climate change on oriental fruit Sutherst RW, Maywald GF, Kriticos DJ (2007) CLIMEX Version 3: User’s Guide. fly in New Zealand and the Pacific. New Zealand Plant Protection, 60, 271–278. Hearne Scientific Software Pty Ltd, Melbourne. www.Hearne.com.au. Kriticos DJ, Watt MS, Potter KJB, Manning LK, Alexander NS, Tallent-Halsell N Taylor S, Kumar L, Reid N, Kriticos DJ (2012) Climate change and the potential (2011) Managing invasive weeds under climate change: considering the current distribution of an invasive shrub, Lantana camara L. Public Library of Science One, and future distribution of Buddleja davidii. Weed Research, 51,85–96. 7, e35565. Kriticos DJ, Webber BL, Leriche A, Ota N, Bathols J, Macadam I, Scott JK (2012) Cli- Troup RS (1983) The Silviculture of Indian Trees. Volume IV. Leguminosae. pp. 345. Mond: global high resolution historical and future scenario climate surfaces for Government of India Press, Nasik, India. bioclimatic modelling. Methods in Ecology and Evolution, 3,53–64. Van der Heijden K (1997) Scenarios: The Art of Strategic Conversation. Wiley, Chiches- Lewandowski I, Clifton-Brown J, Scurlock J, Huisman W (2000) Miscanthus: Euro- ter. pean experience with a novel energy crop. Biomass and Bioenergy, 19, 209–227. Verbruggen A, Al Marchohi M (2010) Views on peak oil and its relation to climate Low T, Booth C (2008) The Weedy Truth About Biofuels. Invasive Speices Council, Mel- change policy. Energy Policy, 38, 5572–5581. bourne. Watt MS, Kriticos DJ, Manning L (2009) The current and future potential distribution Mack RN, Lonsdale WM (2001) Humans as global plant dispersers: getting more of Melaleuca quinquenervia. Weed Research, 49, 381–390. than we bargained for. BioScience, 51,95–102. Watt MS, Kriticos DJ, Lamoureaux SL, Bourdot GW (2011) Climate change and the McConnachie AJ, Strathie LW, Mersie W et al. (2011) Current and potential geo- potential global distribution of the invasive weed Nassella trichotoma. Weed Science, graphical distribution of the invasive plant Parthenium hysterophorus (Astera- 59, 538–545. ceae) in eastern and southern Africa. Weed Research, 51,71–84. Webber BL, Yates CJ, Le Maitre DC et al. (2011) Modelling horses for novel climate Mukta N, Sreevali Y (2009) Investigations of an uncommon accession of Pongamia courses: insights from projecting potential distributions of native and alien Aus- pinnata (L.) Pierre. Indian Forester, 135, 293–295. tralian acacias with correlative and mechanistic models. Diversity and Distribu- Mukta N, Sreevalli Y (2010) Propagation techniques, evaluation and improvement of tions, 17, 978–1000. the biodiesel plant, Pongamia pinnata (L.) Pierre - a review. Industrial Crops and Products, 31,1–12.