Pierceland data. In this case the conventional limits indicate a sipficant increase in the percentage of B. harveyi parasitism, compared to 1955, but the quadratic limits suggest that this is not so. The calculations for obtaining the quadratic limits are lengthy in comparison to conventional methods, but are recommended on the basis of these results. SummaryA technique is presented for estimating parasitism of larch sawfly cocoons by B. harveyi in survey collections. The cocoons are obtained from 4-squarefoot moss filled trays placed under the tree crowns. The trays should be scattered throughout each plot, because of intra-plot variability in parasitism. Twenty trays per plot will provide satisfactory estimates of B. harveyi parasitism. Confidence limits for the proportion parasitized should be estimated by a quadratic approximation, and not by the conventional use of Student's t.

Can. Ent. 101: 785-818 (1969) A generalized competition model for predators or parasites was developed from data obtained from a specific parasite-host system. It was structured in three parts. The first simulates the effects of exploitation, where the number of attacks and their distribution among prey or hosts determine how many prey or hosts survive. Since the negative binomial distribution described these distributions comirrentlp, the exploitation submodel was developed from it. T h e second portion of the competition model concerned interference between searching predators and parasites. Although interference is a universal pl~enomcnon, we were able to show that its effects become important only at predator densities much higher than those that occur in nature. Thus the interference component can be essentially ignored. The third and final component concerned the outcome of competition between parasite progeny within their host. It was developed from Fujii's competition model which allows for the simulation of both scramble and contest types of competition.These three submodels of competition were combined and coupled with a previously published model of the effects of prey density on attack. In this way the full consequences of different prey and predator densities could be simulated using a model whose constituent parts had been carefully rested for descriptive aderluacy. The simulations showed the way individual predator attack, per cent predation, and progeny production were affected by different degees of contagion in the disrribution of attacks, by scramble vs. contest competition, and by the degree to which parasites could avoid hosts already attacked.Volume 101 THE CANADIAN ENTOMOLOGIST 787 attacking the European sawfly, Neodiprion sertifer (Geoff.). Both of these species were originally Palaearctic, but both are now established in eastern North America (Griffiths 196 1 ;Lyons 1964).N . sertifer is a univoltine egg-overwintering sawfly (Griffiths 1959). Cocooned larvae, the stage attacked by P. basizonus, can be obtained in large numbers simply by mass-rearing nearly-mature feeding-stage larvae until spin-up. The cocoons can then be stored at 0°C until needed, with very little mortality and no development.An attacking P. basizonus inserts its ovipositor through the sawfly cocoon, paralyzes the prepupa within, and lays a single egg in the cavity between the prepupa and the cocoon's inner surface (Griffiths 1961). P. basizonus is an ideal

The possibility of imperfect coincidence between the appropriate stages of Neodiprion sertifer (Geoff.) and two of its important parasites was demonstrated. One of the parasites, the indigenous ichneumonid Exenterus canadensis Prov., which attacks late-stage larvae, has good spatial coincidence; but some members of each generation suffer from imperfect temporal coincidence, or asynchrony, caused by the interaction of temperature influence on parasite development rate and temperature variability between development sites in the litter. The second parasite, Pleolophus basizonus (Grav.), is an introduced, multivoltine ichneumonid cocoon parasite. It may be imperfectly synchronized in its first generation each year and may show imperfect spatial coincidence in all generations through its inability to attack host cocoons beneath approximately 1 in. or more of litter.The intricate relations between parasite and host density, time, attack, and coincidence were investigated using the basic functional response submodel developed by Holling, a submodel that describes changes in oviposition behaviour with time, and a submodel that predicts the number of hosts attacked, given the number of eggs laid and data on the distribution of eggs among hosts. In the two species studied, the effect of asynchrony in one generation cannot be considered without considering the influence of superparasitism. At low host densities, superparasitism largely buffers the effects of decreased synchrony. This buffering effect decreases as host density increases until when each parasite is attacking all the hosts it can, it is almost eliminated. Imperfect spatial coincidence in one generation merely lowers the usable host density. Thus its effect can be seen in the functional response of the parasite to host density. When host–parasite interactions over 25 to 35 host generations were simulated, using initial conditions resembling those ensuing when small numbers of both host and parasite invade a previously unattacked stand, populations became stable after passing through one or more oscillations. Decreasing temporal or spatial coincidence increased host and parasite densities at the peaks of oscillations and increased the ultimate steady density of host and parasite, until coincidence was reduced to nearly half. At this level, the host escaped the regulating ability of both species of parasites.

Aims: To examine a range of udder and teat traits in Romney ewes and to describe the frequency with which different scores occur, and to investigate associations between lamb survival to weaning and ewe udder and teat scores. Methods: Mixed-age, mature Romney ewes (n = 1,009) were enrolled from a commercial sheep flock located in the Wellington region of New Zealand in January 2017. A range of udder and teat traits were scored in all ewes, using visual assessment and palpation, at pre-mating (February), pre-lambing (October), docking (November) and weaning (January 2018). During the lambing period each newborn lamb was matched to its dam, with lamb mortalities recorded until weaning. Associations between udder and teat scores and lamb survival to weaning were examined using multivariable models for each udder-scoring time. Results: Records from 981 ewes and 1,822 live-born lambs were included in analyses, with 252 (13.8%) lambs recorded dead between birth and weaning. Lambs born to ewes with pre-mating udder scores of lump or hard had 4.9 (95% CI = 2.6-9.6, p = 0.003) and 3.0 (95% CI = 1.5-6.1, p < 0.001) increased odds of failure to survive to weaning, respectively, compared with lambs whose dams had normal udder scores. Lambs born to ewes with mastitis at docking or weaning had 3.0 (95% CI = 1.5-5.9, p = 0.001) and 3.9 (95% CI = 1.3-11.6, p = 0.013) increased odds of failure to survive to weaning, respectively, compared with lambs whose dams did not have mastitis. Offspring of dams with asymmetrical udders at docking or weaning had 3.3 (95% CI = 2.2-4.9, p < 0.001) and 2.5 (95% CI = 1.5-4.0, p < 0.001) increased odds of failure to survive, respectively, compared with lambs whose dams had symmetrical udders. Conclusion and Clinical Relevance: Pre-mating udder palpation scores of hard or lump were associated with increased odds of lambs not surviving to weaning compared with normal scores, and could be used to identify ewes that are likely to be unsuitable for retaining in the breeding flock. Farmers could also use clinical mastitis scores and udder symmetry scores at docking or weaning to identify ewes whose lambs had greater odds of failure to survive to weaning. However these scores do not provide an indication of future performance, therefore further investigation into the impact of the present season's score on future seasons' lamb survival is required.

No abstract