Cat’s claw creeper vine, Dolichandra unguis-cati (L.) Lohmann (syn. Macfadyena unguis-cati (L.) Gentry) (Bignoniaceae), is a major environmental weed in Australia. Two distinct forms of this weed (‘long’ and ‘short’ pod), with differences in leaf morphology and fruit size, occur in Australia. The long pod form has only been reported in less than fifteen localities in the whole of south-east Queensland, while the short pod form is widely distributed in Queensland and New South Wales. This study sought to compare growth traits such as specific leaf area, relative growth rate, stem length, shoot/root ratio, tuber biomass and branching architecture between these forms. These traits were monitored under glasshouse conditions over a period of 18 months. Short pod exhibited higher values of relative growth rates, stem length, number of tubers and specific leaf area than long pod, but only after 10 months of plant growth. Prior to this, long and short pod did not differ significantly. Higher values for these traits have been described as characteristics of successful colonizers. Results from this study could partly explain why the short pod form is more widely distributed in Australia while long pod is confined to a few localities.

Cat’s claw creeper, invasive species, competitiveness, relative growth rate, successful colonizers, traits, biomass, tubers

Invasive plant species continue to threaten biodiversity and ecosystem function globally (Heckel 2004; Pimentel et al. 2005). A fundamental objective of invasion ecology is to identify a suite of plant traits that may determine invasion success in novel environments (Pyšek and Richardson 2007; Richardson and Pyšek 2006; van Kleunen et al. 2010). An outcome of this search can be traced back to Baker’s ideal weed hypothesis, in which Baker (1965) proposed a set of plant traits most likely to be exhibited by invasive species . Comparative studies between exotic invasive species and their native non-invasive congeners have contributed immensely to our understanding of traits that promote colonisation and invasion success by some species (e.g. van Kleunen et al. 2011).

It has proven difficult to consistently find a correlation of the same set of traits with invasiveness, likely because of the varying effects of environmental factors on different plant species (Alpert et al. 2000; Burns 2006). Studies have shown that no particular trait solely confers invasiveness on a species, rather it is how a species responds to different environmental conditions that contributes to its fitness and abundance (Firn et al. 2012; Leishman et al. 2010; Osunkoya et al. 2010; Pattison et al. 1998). Plastic responses of invasive plants to varying environmental conditions increase their competitiveness and fitness (Claridge and Franklin 2002). Therefore, multiple factors likely explain the success of invasive plant species (Blumenthal 2005; Daehler 2003; Lamarque et al. 2011; Leffler et al. 2014; Leung et al. 2004; MacDougall et al. 2009). For example, Burns (2006) found that invasive species had higher specific leaf area (SLA) and relative growth rates (RGR), but only under certain environmental conditions. Nevertheless, a pattern of relatedness to invasiveness has been reported for some plant traits (Pyšek and Richardson 2007). Mostly, traits that have direct relatedness to plant physiological performance such as leaf area ratio, growth rate, shoot/root allocation and propagule pressure show marked differences between evidently invasive and non-invasive species (Grotkopp et al. 2002; van Kleunen et al. 2010).

Invasive species were shown to have higher values of traits like SLA (Burns 2006; Lake and Leishman 2004), RGR (Dawson et al. 2011), and more biomass allocated to organs like stems, resulting in taller plants (Gallagher et al. 2015; Stanisci et al. 2010; van Kleunen et al. 2015). High SLA is often associated with high RGR (Grotkopp and Rejmánek 2007), although other studies have not found that trend (see, for example, Garcia-Serrano et al. 2005). Overall, fast growing plants have generally been found to be more likely to be invasive than others (Blumenthal and Hufbauer 2007; Lake and Leishman 2004; Richardson 1998). Higher values for these traits in invasive species compared to less invasive ones imply different strategies for capturing and using resources such as light, carbon, nitrogen and moisture (Gallagher et al. 2015). Because resources are almost always limiting in the environment (Cordell et al. 1998), efficient use of limiting resources by invasive species can enhance their colonizing success (Pattison et al. 1998). In disturbed environments, species that are better able to exploit fluctuating resources will likely invade the system (Cordell et al. 1998; Leffler et al. 2014; van Kleunen et al. 2010).

Most studies aimed at understanding differences in traits associated with invasion success have used native species as control plants (Muth and Pigliucci 2006). The limitation of this approach is that these native species may already be invasive elsewhere (van Kleunen et al. 2010). For example, some native species used in a comparative study by Godoy et al. (2011) were reported to be invasive in other parts of the world. Other studies have also shown that these traits do not always differ between invasive and non-invasive species (Meiners 2007; Smith and Knapp 2001; Thompson et al. 1995). An assessment of 122 species including non-native invasive and native species that occupy disturbed areas did not find significant differences in these traits (Leishman et al. 2010). Muth and Pigliucci (2006) argue that some native species were shown to have invasive tendencies in their native range, implying that introduced vs native species comparisons may not always be informative (but see Blossey and Notzold 1995; Callaway and Ridenour 2004; Dawson et al. 2015; Keane and Crawley 2002; van Kleunen et al. 2011). There could also be a bias in choosing highly competitive invasive species and comparing them with known weak native competitors in pairwise experiments (Vila and Weiner 2004) or comparing phylogenetically non-related species (Burns 2006).

Our understanding of invasiveness traits could be better enhanced by comparing related non-native species of varying levels of colonization success (Kolar and Lodge 2001; Muth and Pigliucci 2006; van Kleunen et al. 2010). In this study, we compare different traits between two forms of an invasive vine, cat’s claw creeper, that appear to have significantly different levels of invasion success. Cat’s claw creeper, Dolichandra unguis-cati (L.) Lohmann (syn. Macfadyena unguis-cati (L.) Gentry) was introduced as an ornamental into Australia from South America in the 1800s (Dhileepan 2012; Downey and Turnbull 2007; Gentry 1976). D. unguis-cati is now a declared environmental weed and considered formally as a Weed of National Significance (WoNS) in Australia (Thorp and Lynch 2000).

Dolichandra unguis-cati prefers forested and riparian habitats, although it also grows vigorously on dry road side sunny environments. It also appears to thrive in most soil types, tolerating a wide range of soil pH (Downey and Turnbull 2007). Two forms of this species with distinct leaf morphology occur in Australia (Dhileepan 2012; Shortus and Dhileepan 2011). The two forms of D. unguis-cati were named long pod (LP) and short pod (SP) due to differences in their average fruit (pod) length at maturity (LP: 700.2 ± 23.5 mm; SP: 300.9 ± 89.6 mm) (Shortus and Dhileepan 2011). While LP occurs in isolated localities of south-east Queensland (Qld), SP occurs extensively in Qld and New South Wales, often in dense infestations (Dhileepan 2012; Downey and Turnbull 2007). These two forms appear to prefer similar habitats, although there is general lack of research on the ecology of this species (Osunkoya et al. 2009). The LP and SP forms have been shown to carry an average of 120 ± 10 and 60 ± 23 seeds per pod at maturity, respectively (Shortus and Dhileepan 2011). Seeds of both forms are two-winged, papery and flattened/oblong in shape, 10–18 mm long, 4.2–5.8 mm wide. The average seed biomass is not significantly different between the forms of D. unguis-cati (mean seed biomass for LP: 16.60 ± 0.65 mg and for SP: 15.65 ± 0.83 mg) (Shortus and Dhileepan 2011). Previous studies have found that the two forms showed differences in some life history traits. Boyne et al. (2013) found a wide variety of leaf morphology for this species, but also reported that SP had significantly more simple leaves than LP.

In a field experiment using plants generated from tuberlings, Taylor and Dhileepan (2012) found that LP produced greater total dry mass (hence higher RGR) than SP although the study did not measure such parameters as specific leaf area (SLA) and leaf area ratio (LAR). SP was shown to have rapid and higher germination rates than LP at varying temperature regimes (Buru et al. 2014). SP was also reported to exhibit significantly higher frequencies of polyembryony than LP, at times one seed producing quadruplet seedlings (Buru et al. 2016). The only study on the seed bank ecology of the most prevalent form (SP) by Vivian-Smith and Panetta (2004), found it to have low seed longevity, usually less than 12 and 1% at 1 year for soil-surface (< 1 cm depth) and 5 cm depth buried seeds, respectively. Osunkoya et al. (2009) also noted some differences in stem density of genets and ramets between the two forms in field samples, but decried lack of data on growth rates and reproductive capacity for the two forms.

Herbarium records and field surveys suggest that LP is widely distributed in the native range, occurring from Mexico, Nicaragua, Costa Rica, Columbia to Brazil, whereas SP appear to be restricted to Paraguay (Dhileepan 2012; K. Dhileepan, personal observations). In Australia, previous field surveys have revealed that there were seven sites in south-east Queensland (Qld) where LP has been reported, two at which it co-occurs with SP (Boyne et al. 2013; Dhileepan 2012; Shortus and Dhileepan 2011). Recently, seven more sites were identified, bringing the total number of known sites to 14 in south-east Qld where LP occurs (Liz Snow (Biosecurity Queensland), pers. comm. 7/03/2016).

The cause for the observed differences in abundance levels between LP and SP is not yet established, but introduction pressure may be one explanation. Reconstructing the invasion history of this exotic species (or the two forms) is not possible because there are no records of their introduction, except that the species was first reported in a Melbourne Nursery catalogue in 1865 (Downey and Turnbull 2007). Introduction history of most ornamental plants is generally not or poorly recorded (Harris et al. 2007; Prentis et al. 2009). Studies on whether there has been any deliberate breeding selection of the species that resulted in the two forms are yet to be done.

Another explanation could be differences in growth strategies between LP and SP. Rapid growth and efficient resource allocation enhance success in colonization, especially during the early stages of plant life history (Bachmann et al. 2012; Luo et al. 2015). Considering that LP and SP show marked abundance differences in Australia, comparing important functional traits of the two forms may assist with understanding whether different growth strategies explain the different populations. Significantly higher values of growth related traits for one form could infer different strategies of resource use (Dawson et al. 2011; Godoy et al. 2011). Here we sought to compare traits such as SLA, RGR, stem length, shoot/root ratio, tuber biomass and branching architecture between the two forms of D. unguis-cati plants grown from seeds. We did this to develop a type of prospectus on the growing strategies of the two forms of D. unguis-cati that may begin to explain differences in their distributions and abundance.

The SP form, which is more widely distributed within eastern Australia, showed faster growing strategies. Higher values of RGR, stem length, number of tubers, and SLA are often indicators of successful colonizers (Gallagher et al. 2015; Holaday et al. 2015). Higher values of RGR normally correlate with high values of leaf area ratio (LAR) and SLA (Garcia-Serrano et al. 2005). These results are in accordance with the predictions of the ‘leaf economic spectrum’ (LES) hypothesis (Wright et al. 2004), which suggests a fundamental trade-off in the traits held by fast- and slow-growing plant species. According to the LES theory, where a species can be found within the spectrum is associated with strategies for resource capture and use. At one extreme are faster growing and highly productive species while on the other end slower growing and more conservative species occupy (Holaday et al. 2015).

Recent evidence, however, suggests the same carbon assimilation strategies are used by invasive and non-invasive plants (Leishman et al. 2010), but invasive plants have a tendency to cluster towards the ‘high return on investment’ end of the world wide leaf economic spectrum (Funk et al. 2013). Although SP seems to lean towards this end of the spectrum for some traits at 18 months, there were significant overlaps with LP earlier in the plants’ growth. Most studies simply consider ‘adult’ traits (e.g. Bachmann et al. 2012; Burns 2006; Hulshof and Swenson 2010), so we know very little about younger plants (but see Luo et al. 2015). In this study, there is evidence that trait differences are minimal up to 10 months old, but after this age our results suggest that they begin to differ between LP and SP. In our study, PCA shows that the two forms are different at 18 months with the variation mostly explained by growth related traits (PC1), followed by difference in how biomass is allocated below- and above-ground (PC2).

Our results also seem to contradict findings by Taylor and Dhileepan (2012) who observed that LP had higher growth rates than SP in the field. These differences could be attributable to environmental (Evans and Hughes 1961) and growing conditions (field vs glasshouse) (Limpens et al. 2012). Moreover, whilst we generated experimental plants from seeds (seedlings) in our experiments, Taylor and Dhileepan (2012) used plants grown from tuberlings. Also, in the current experiment, plants were not supported while in Taylor and Dhileepan (2012) they were supported with trellises. Our study could also be limited by lack of additional nutrients in the commercial potting mix, although all individuals in the experiment were treated the same and therefore growth and response is comparable.

Although SP had slightly higher values of SLA, it had lower values of LAR when compared to LP. Because LAR is a measure of the leafiness of a plant (Radford 1967), our results imply that although LP might be leafier, SP invests more biomass to branches and stems, which could be a benefit for growing taller and spreading wider. Higher SLA has been positively correlated with high RGR and more rapid turnover of leaf material (Grotkopp et al. 2002). By rapid growth and quick tissue turnover, plants ensure that they outcompete others for limited resources (Gallagher et al. 2015). High growth rates by more successful species are particularly important in the seedling stage of a plant’s life history (Grotkopp et al. 2002). Developing more branching is highly advantageous for vines as it is a way to increase LAR and LMR for maximum harvesting of light in order to optimise photosynthesis. Our results partly corroborate this hypothesis as we found that SP displayed higher values for SLA and LMR (but not for LAR) than LP. By developing more branches than LP (indicated by higher ADI values), SP can effectively out-compete other competitors in the environment for limiting resources.

Transformer plants such as vines like D. unguis-cati, thrive in growing vertically and spreading horizontally to monopolise light environments (Heckel 2004). The negative impacts of this group of plants lie in their ability to smother host tree canopies that they use as supporting structures (Harris and Gallagher 2011; Harris et al. 2007; Zhang et al. 2004). Dolichandra unguis-cati forms thick mats of intertwining creeping stems and branches on forest floors (Osunkoya et al. 2009). Thus, ensuring rapid elongation of stems and a higher branching architecture may be central to the successful colonization of empty habitats by SP. This pattern of growth reduces light availability to low lying vegetation and may prevent recruitment of native plants (Downey and Turnbull 2007; Schnitzer and Bongers 2002; Zhang et al. 2004).

This study shows that SP develops subterranean tubers early in its development while LP seems to delay tuber development. Tubers are used as a sink or storage organs for moisture and photo-assimilates and they may also regenerate producing new plants (Janeček and Klimešová 2014; Orthen 2001; Schubert and Feuerle 1997). Apart from seed germination (Buru et al. 2014; Vivian-Smith and Panetta 2004), D. unguis-cati propagates vegetatively through tubers (Downey and Turnbull 2007; Osunkoya et al. 2009). Horizontal stems and branches trailing along the ground develop roots at nodes, which in turn develop tubers. If the new plants regenerating at the nodal tubers are severed from the mother plant, they grow independently as genets. This study shows that SP develops significantly more tubers per plant than LP, which could be a clonal survival strategy to increase its competitiveness. Clonal growth of a species may enhance its invasion success by way of rapid formation of monocultures (Aguilera et al. 2010; Pyšek and Richardson 2007). Liu et al. (2006) found a positive relationship between clonality and invasiveness. They found that more than 66% of the most invasive plants they studied in China were clonal. Resource storage by clonal plants function as a back-up measure in case of adverse alterations in the growth conditions of the plant (Suzuki and Stuefer 1999). Tubers can also remain dormant for extended periods belowground as a stress tolerance strategy (Orthen 2001).

Previous studies have shown SP to exhibit more rapid and higher germination rates than LP at various temperatures (Buru et al. 2014) and a higher frequency of polyembryony than LP (Buru et al. 2016). Seeds of the two forms do not differ in their average mass (Shortus and Dhileepan 2011). This study has shown that SP displayed superior values of traits known to be associated with successful invaders (Chun et al. 2007; van Kleunen et al. 2015). Therefore it may be safe to assume that were the two forms to be introduced into novel environments at the same time, SP would likely be more successful in colonizing the habitats than LP (Gallagher et al. 2015; Godoy et al. 2012; Kolar and Lodge 2001; Pyšek and Richardson 2007; van Kleunen et al. 2010). Thus, our results partly explain why SP seems to be abundant in Australia, although LP is postulated to also have a potential to become widespread if not carefully managed (see Taylor and Dhileepan 2012).

We acknowledge both the Government of Botswana and the Department of Agriculture and Fisheries (Queensland, Australia) for part-sponsorship of this study through student scholarships to J.C Buru. We thank members of the Plant Structure and Systematics Research Group and Dr. Tanya Scharaschkin for providing valuable feedback on earlier versions of the manuscript. We also thank three reviewers who gave very useful comments that helped improve the manuscript greatly.