DAVID A. NORTON. JENNY J. LADLEY. Conservation ..... We thank Dave Kelly, Peter de Lange, and Nick Reid for comments on a draft manuscript; the ...
New Zealand Journal of Botany, 1997, Vol. 35: 337-343 0028-825X/97/3503-0337 $7.00 © The Royal Society of New Zealand 1997
337
Development of non-destructive age indices for three New Zealand loranthaceous mistletoes
DAVID A. NORTON JENNY J. LADLEY Conservation Research Group School of Forestry University of Canterbury Private Bag 4800 Christchurch, New Zealand
INTRODUCTION
While it is possible to obtain accurate information on the age of many woody plants by counting the number of annual rings (Norton & Ogden 1987), estimation of age using this technique is difficult with mistletoes (Loranthaceae and Viscaceae). This may be so because the mistletoe or its host does not produce annual growth rings (e.g., Patel 1991; Reid et al. 1995) or for conservation reasons it is not conASHLEY D. SPARROW sidered desirable to destructively sample the mistleDepartment of Plant and Microbial Sciences toe in order to age the haustorial connection (Dawson University of Canterbury et al. 1990a). Accurate information on plant age is Private Bag 4800 essential if we are to understand better the populaChristchurch, New Zealand tion dynamics of mistletoes. Furthermore, it has been suggested that many aspects of mistletoe biology are Abstract We investigated a variety of non-de- age-dependent (Schulze & Ehleringer 1984; Dawson structive measures as potential predictors of mistle- et al. 1990b; Powell & Norton 1994). Several authors have shown that it is possible to toe age as determined anatomically for three mistletoe species, Alepis flavida, Ileostylus age mistletoes anatomically by counting the number micranthus, and Tupeia antarctica. We show that the of growth rings laid down by the host plant since the diameter of the host stem immediately below the haustorial connection was established (Srivastava & haustorial attachment is consistently the best predic- Esau 1961; Calvin 1967; see also Menzies 1954). tor of mistletoe age with R2 values of 0.622-0.849. However, this may not always be possible or desirWe suggest that host branch diameter can be used able. Work by Dawson et al. (1990a) with the in future studies of mistletoe population dynamics viscaceous mistletoe Phoradendron juniperinum in and other age-dependent aspects of mistletoe ecol- western North America has shown that the number ogy without destructively sampling mistletoe of bifurcating branching events on the longest stem populations as it provides a good indication of mis- of the mistletoe was strongly correlated with mistletletoe age for the mistletoe-host pairs we studied. toe age based on anatomical work. For the AustralHowever, these relationships have been derived from ian loranthaceous mistletoes Amyema quandang and single sites and hosts for each mistletoe species, Lysiana exocarpi, Reid & Lange (1988) argued that suggesting caution when applying them at other sites the maximum diameter of the host branch proximal to the haustorium was proportional to mistletoe age or to other hosts. because seedlings of most mistletoes established on Keywords Alepis flavida; Ileostylus micranthus; young host branches. However, because both their Tupeia antarctica; Loranthaceae; mistletoe; ageing mistletoes and host species did not produce clear annual growth rings, they were not able to corroborate this. Reid et al. (1995) suggested that regular vegetative growth and branching patterns (cf. Dawson et al. 1990a) can be used as an index of mistletoe age in mistletoe species that have an annual growing season. B96067 Evidence suggests that, at least in some parts of Received 5 November 1996; accepted 3 April 1997 New Zealand, loranthaceous mistletoes (Alepis
New Zealand Journal of Botany, 1997, Vol. 35
338
A
•
„ s •
B
S S
10
D »
o
20
30
40
Fig. 1 The relationship between mistletoe age as determined by anatomical analysis and a number of non-destructive potential indices of age for Alepis jlavida. The solid line is the fitted regression, with 95% confidence intervals indicated by the dotted lines (n = 26). A, Host branch diameter (mm); B, Mistletoe basal diameter (mm); C, Mistletoe volume (cm3); D, Resting bud scar number; E, Maximum number of branch orders.
.
E, 900-1000 m a.s.l., Craigieburn Ecological District) where it is common in monospecific Nothofagus solandri forest; Ileostylus micranthus from scattered remnant Podocarpus totara trees in farmland near Wakefield, south of Nelson (41° 25' S, 173° 02' E, 60 m a.s.l., Moutere Ecological District); and Tupeia antarctica from roadside Chamaecytisus palmensis shrubs at Wainui on Banks Peninsula (43° 49' S, 172° 54' E, 10 m a.s.l., flavida, Ileostylus micranthus, Peraxilla colensoi, Akaroa Ecological District). Peraxilla tetrapetala, Tupeia antarcticd) are less At each site we aimed to sample 30 mistletoe abundant now than they have been in the past and plants spanning the full range of mistletoe plant sizes there is considerable concern about their future abun- present. Where possible, we attempted to sample dance (papers in de Lange & Norton 1997; Norton only one mistletoe plant per host plant to reduce & Reid 1997). Reflecting these concerns, all of New impacts on hosts. However, at the Wainui site where Zealand's loranthaceous mistletoes have been the host plant is a naturalised shrub, we relaxed this ranked as local or threatened (Cameron et al. 1995). constraint. Plants were sampled by cutting the misBecause of this, destructive methods of aging mis- tletoe and its host branch off the host tree. For each tletoes are considered unacceptable and permission mistletoe, measurements were made of the diameter for this is unlikely to be granted on lands managed of the mistletoe stem immediately above the hausby the New Zealand Department of Conservation (P. torium, the diameter of the host branch immediately J. de Lange pers. comm.). In this study, we sought below the haustorium, the length of the longest axis to develop non-destructive methods to age New through the mistletoe and of two axes perpendicuZealand's loranthaceous mistletoes as a basis for lar to this (from which we calculated the volume of better understanding their population dynamics and the mistletoe based on the formula for an ovoid), the other aspects of their ecology. maximum number of annual resting bud scars present between the haustorium and branch tip on the mistletoe, and the maximum number of branch orders present on the mistletoe. METHODS The mistletoe was then sectioned at the point of We sampled three mistletoe species at sites in north- the haustorial attachment so that the age of the misern South Island, New Zealand: Alepis flavida in tletoe could be determined anatomically based on the Craigieburn Conservation Park (43° 09' S, 171° 43' number of annual growth rings formed by the host
Norton et al.—Non-destructive age indices for mistletoes
339
Fig. 2 The relationship between mistletoe age as determined by anatomical analysis and a number of non-destructive potential indices of age for Ileostylus micranthus. The solid line is the fitted regression, with 95% confidence intervals indicated by the dotted lines (n = 30). A, Host branch diameter (mm); B, Mistletoe basal diameter (mm); C, Mistletoe volume (cm3); D, Resting bud scar number; E, Maximum number of branch orders.
B
10
15
20
D
0
2
4
8
10
12
before analysis. Thus, the regressions for any combination of age and its estimators are in fact curvilinear (logarithmic or power curves). The extent of non-linearity was tested by fitting a power curve to each data set (logarithmic or power) and testing the Chamaecytisus palmensis, growth rings were at estimate of slope (which corresponds to the power times indistinct and age estimates for this species exponent) against a linearly expected value of one may be less accurate than for the other two. Although using a simple t-test (t = (slope-l)/SEslope). The fitNew Zealand mistletoes show no winter growth ted equations and their confidence intervals were (Powell & Norton 1994), growth rings are indistinct back-transformed for the purposes of graphing and in the three mistletoe species studied here (Patel estimation. 1991) and could not be used to age the mistletoes. For each of the three mistletoe species simple linear regression (using S-PLUS software; MathSoft 1995) was used to test for the correlation of anatomi- RESULTS cally estimated age with each of the five non-destruc- Predictions of anatomical mistletoe age using all five tive measures. In addition, a multiple regression was non-destructive predictors of mistletoe age are sigcalculated for age against all five measures (for each nificant for all three mistletoe species, although the species). Before fitting the regressions, age and all variance explained in the regressions shows considfive potential estimators of age were tested for the erable variation (Table 1, Fig. 1-3). Of the three regression's assumption of normality. As a result of mistletoe species, Alepis flavida age is best modelled these tests, age, host branch diameter, mistletoe di- by the non-destructive predictors with R2 values ameter, and mistletoe volume were log-transformed ranging from 0.686-0.849 (Table 1, Fig. 1). plant after the haustorium had formed (cf. Dawson et al. 1990a). For Nothofagus solandri and Podocarpus totara, growth ring formation is known to be annual (Wells 1972; Norton 1984) and growth rings are distinct and easily counted. However, for
New Zealand Journal of Botany, 1997, Vol. 35
340
Fig. 3 The relationship between mistletoe age as determined by anatomical analysis and a number of non-destructive potential indices of age for Tupeia antarctica. The solid line is the fitted regression, with 95% confidence intervals indicated by the dotted lines (n = 30). A, Host branch diameter (mm); B, Mistletoe basal diameter (mm); C, Mistletoe volume (cm3); D, Resting bud scar number; E, Maximum number of branch orders.
B
20
40
60
D
0
20
Table 1 Summary of the fit of the regressions (R2, P(regression)) for predicting mistletoe age from non-destructive measurements for the three mistletoe species and results for test for non-linear regressions (P(nonlinear)). Predictor
R2
Alepis flavida (n = 26) Host branch diameter (log) 0.849 Mistletoe basal diameter (log) 0.773 Mistletoe volume (log) 0.796 Max. number of annual resting bud scars 0.719 0.686 Max. number of branch orders 0.866 Multiple regression (all predictors) Ileostylus micranthus (n = 30) Host branch diameter (log) 0.790 Mistletoe basal diameter (log) 0.602 Mistletoe volume (log) 0.699 Max. number of annual resting bud scars 0.653 0.545 Max. number of branch orders 0.834 Multiple regression (all predictors) Tupeia antarctica (« = 30) Host branch diameter (log) 0.622 Mistletoe basal diameter (log) 0.295 Mistletoe volume (log) 0.352 Max. number of annual resting bud scars 0.445 Max. number of branch orders 0.384 Multiple regression (all predictors) 0.688
P (regression) P (non-linear) O.0001
